Article Text
PDF
Abstract
Objectives To describe the range of concurrent cardiac malformations in biliary atresia (BA) while providing a functional framework of risk.
Methods Demographic and variables were collected from a prospectively maintained single-centre database. Infants were grouped according to a cardiac functional framework (A=acyanotic, B=cyanotic and C=insignificant shunt). Primary outcome was set as clearance of jaundice (bilirubin ≤20 μmol/L) following Kasai portoenterostomy (KPE). Native liver survival and overall actuarial survival were compared with a date-matched control infant with BA (n=77). P value <0.05 was regarded as significant.
Results 524 infants with histologically confirmed BA were treated between January 1999 and December 2018, 37 (7%) had a concurrent cardiac anomaly (A: n=23 (62%), B: n=10 (27%), C: n=4 (11%)). Infants with biliary atresia splenic malformation (BASM) or cat-eye syndrome (CES) contributed over half of the cases (21/37; 57%).
Overall, 20 (54%) infants cleared jaundice (vs 50/77 (65%) controls; p=0.2), but with higher mortality compared with the non-cardiac controls (15/37 (40%) vs 3/77 (4%); HR 15.5 (95% CI 5.5 to 43.4); p<0.00001). Infants requiring cardiac intervention in the first year of life (n=15) were more likely to clear jaundice (6/7 vs 2/8; p=0.04) and had a trend towards higher survival (6/7 vs 3/8; p=0.1) when KPE followed cardiac surgery. Yet, the type of cardiac pathology did not impact clearance of jaundice or mortality.
Conclusion We propose the term cardiac-associated biliary atresia (CABA) as a high-risk group. We believe that restorative cardiac surgery should precede KPE wherever possible to improve outcome.
- cardiac surgery
- congenital abnorm
- paediatric surgery
- hepatology
Statistics from Altmetric.com
What is already known on this topic?
Biliary atresia, particularly syndromic biliary atresia (eg, biliary atresia splenic malformation and cat eye syndrome), are known associations with congenital cardiac anomalies.
This subgroup of biliary atresia is at particular risk of death within the first year.
What this study adds?
We define the entity cardiac-associated biliary atresia (CABA) an important prognostic subgroup and the leading cause of BA mortality in Europe and North America.
We provide a functional framework of risk in the CABA group.
We suggest a ‘heart-first’ strategy to improve both liver outcome and overall survival.
Introduction
Biliary atresia (BA) is an infantile obliterative cholangiopathy that presents with persistent conjugated hyperbilirubinaemia, white stools and dark urine leading to, if untreated, liver fibrosis and cirrhosis. Although its aetiology remains speculative in most cases, it is clear that there are a number of different variants that can be clinically, occasionally genetically, characterised. There are at least two syndromic associations with therefore developmental origins—biliary atresia splenic malformation (BASM) and cat-eye syndrome (CES).1–3 Alternatively, others may be prompted by perinatal viral infection with cytomegalovirus IgM-positive BA being most clearly defined.4 5
Associated cardiac anomalies have long been known to feature in both syndromic types although the exact anomaly is not stereotypical unlike the typical visceral anomalies seen in BASM. The incidence of those with associated cardiac anomalies varies from 3.4% to to 16% in eastern and western BA series.6–11
Our aim was to characterise the subset of infants from the perspective of their cardiac anomalies—cardiac-associated biliary atresia (CABA). We sought both to define a functional clinical framework and a strategy to better manage these infants.
Patients and methods
King’s College Hospital, London is the largest of the three specialist paediatric liver centres in the UK treating infants from London and South-East England predominantly. Clinical and demographic features including outcome were determined from a prospectively maintained database for the period January 1998 to December 2018. Kasai portoenterostomy (KPE) was performed by a single surgeon (MD) and liver transplantation was considered in all infants where necessary. The postoperative regimens varied during the period and in particular high-dose steroids were introduced from 2005 and adjuvant antiviral therapy was available from 2015.
Precise diagnosis of a detected cardiac anomaly was made in one of three London paediatric cardiac units in all cases (Evelina London Children’s Hospital, Great Ormond Street Hospital and The Royal Brompton Hospital). Contemporaneous echocardiography and/or cardiac angiography (where available) was reviewed retrospectively to define the anomalies for this analysis (KP, OM) and assign each infant to one of three functional groups (prior to any intervention). Group A were defined as acyanotic anomalies expressing at least one of three features: left-to-right shunt; obstructive lesions of the left heart (eg, coarctation of the aorta) or a cardiomyopathy. Group B were defined as cyanotic anomalies including right heart obstruction (eg, tetralogy of Fallot) and infants with a duct-dependent pulmonary circulation. Infants in group C were defined as having non-haemodynamically significant pathology or pulmonary hypertension.
The primary measure of outcome post-KPE was defined as clearance of jaundice (≤20 μmol/L bilirubin) at any point within 6 months. Secondary outcomes included native liver survival (NLS) and overall survival (OS). A date-matched group of infants with non-cardiac BA were used as comparative controls on the basis of 2:1 ratio.
This study was registered as a service evaluation of clinical practice (registration no. CH043) and therefore did not need formal institutional approval by a research ethics committee.
Data are presented as median (range), unless otherwise stated. Categorical data were compared using χ2 or Fisher’s exact test, and ordinal data were compared with a non-parametric Mann-Whitney U test. Kaplan-Meier actuarial NLS and OS curves were compared using a log-rank test. A p value of <0.05 was regarded as statistically significant. All statistical analysis was performed using GraphPad Prism (V.5.00 for Windows; GraphPad Software, San Diego, CA, USA, www.graphpad.com).
Results
Demography
A total of 524 infants with histologically confirmed BA were managed during the 20-year period, from which 37 (7%) infants with concurrent congenital cardiac anomalies were identified (CABA) and 77 were defined as date-matched controls. CABA was subclassified as group A (n=23), group B (n=10) and group C (n=4). Infants with CABA tended to be female predominant (25 (67%) vs 38 (49%) controls; p=0.06) and more likely to have had abnormal antenatal ultrasound scans (13/37 (35%) vs 6/77 (8%); p<0.001), where cardiac abnormality was suspected on seven of the antenatal scans in the CABA group. The presence of gestational diabetes was comparable between CABA and control groups (1/37 vs 6/77, p=0.4). Over 80% (30/37) showed further non-cardiac syndromic anomalies including BASM (n=18) and CES (n=4), and in the other non-syndromic (7/16; 44%) infants showed an extra non-cardiac anomaly(s) including malrotation, other atresia and ectopic ureter. Chromosomal abnormalities were identified in five (13.5%) infants of the CABA group (chromosome 22 abnormalities compatible with CES (n=4) and a ring chromosome 18 (n=1)), while one infant in the control group had a mosaic chromosome 22 (p=0.01).
Three cases were referred from overseas, while the remainder were born in the UK. Further demographic data and associated anomalies of the CABA group are shown in table 1.
Demographics of cardiac-associated biliary atresia (CABA) group in comparison with the control group
Cardiac anomalies
Congenital cardiac anomalies ranged from atrial septal defect to an extremely complex restrictive left heart anomaly with aberrant pulmonary venous drainage not amenable to surgical correction (table 2). Twenty-one (57%) infants underwent ≥1 cardiac intervention(s), with no further planned intervention in any of the surviving patients at the time of their last cardiology follow-up. Nearly two-thirds of the infants were classified as group A; nonetheless, there was no statistical difference in the distribution of BASM among the groups (table 2).
Outcomes and cardiac anomalies in CABA subgroups
Outcome
Overall, 34 (92%) infants underwent KPE at 55 (23–261) days (vs 52 days in controls; p=0.77). Two infants in group B (tetralogy of Fallot and total anomalous pulmonary venous return) were found to have biliary agenesis at time of laparotomy and were listed for primary liver transplantation. A third, with pulmonary stenosis, was also listed for primary liver transplant after a laparotomy showed type 3 BA and cirrhosis.
There was no difference in the proportion that cleared their jaundice following KPE (20/34 (59%) vs 50/76 (66%); p=0.23). Similarly, native liver survival was comparable with controls (HR 1.2 (95% CI 0.69 to 2.1), p=0.49) (figure 1A). Liver transplantation was performed in eight (22%) infants with CABA, while two listed infants died of cardiac-related pathology. Two further infants were denied listing for orthotopic liver transplantation due to uncorrectable complex cardiac lesions and unfavourable prognosis, and both died later of cardiac-related causes. Characteristics and outcome comparison of the CABA and control group are shown in table 3.
Native liver survival (A) and overall survival (B) in cardiac-associated biliary atresia (CABA) and controls.
Outcomes of CABA and control group
There was a 10-fold difference in mortality in the CABA group (15/37 (40%) vs 3/77 (4%); HR 15.5 (95% CI 5.5 to 43.4); p<0.0001) (figure 1B). Children with CABA died at 0.5 (0.2–10.4) years, and of these 73% occurred in the first year of life.
The type of cardiac pathology did not impact clearance of jaundice post-KPE (group A: 11/22, group B: 5/8 and group C: 4/4, p=0.68). Similarly, we did not detect a difference in mortality between the cardiac groups (group A: 9/23, group B: 5/10 and group C: 1/4, p=0.6) (table 2).
Effect of timing of KPE
In order to test the significance of the relative timing of cardiac and KPE surgeries on outcomes, we analysed patients with CABA who required cardiac surgery in the first year of life. Those defined infants who require early cardiac intervention are those with a serious heart defect with unstable circulation that requires rapid attention.12 Fifteen (40%) infants with BA underwent cardiac surgery for a significant cardiac lesion in the first year of life; eight had their KPE before this intervention. The median age at first cardiac surgery was at a significantly older age when KPE preceded cardiac surgery (181 vs 17 days; p=0.003); however, the median age at KPE was comparable (50 days vs 57 days; p=0.3).
The number of infants that cleared jaundice following KPE was significantly reduced in those who had KPE-first (2/8 vs 6/7; p=0.04). Similarly, mortality was considerably higher in the KPE-first group; 5/8 (62.5%) compared with the KPE-after group 1/7 (14%). Yet, the distribution of cyanotic cardiac lesions was similar between the two groups (3/8 vs 4/7; p=0.6). The demographics and outcomes of the two groups are shown in table 4.
Outcomes of BA in relation to timing of KPE
Discussion
This appears to be the first study to characterise infants with CABA and stratify them on the basis of risk.
The incidence of congenital cardiac anomalies in the UK is about 6.8 per 1000 births13 with the incidence of cardiac anomalies in BA reported as 8%–16% from multicentre North American reviews.9–11 Our estimate of 7% is therefore slightly lower. A recent large multi-centre Chinese study6 quoted associated cardiac anomalies in only 37/851 (4.3%) patients with BA. Similarly, two smaller single-centre series from Japan7 and India8 quote smaller figures (6.8% and 3.4%, respectively), and this may reflect the diminished proportion of BASM seen in Asian series. Of note, the range of cardiac anomalies we encountered was far more complex that what was reported in the Chinese series.6
Cardiac anomalies are a key feature of both BASM1 2 and CES,3 and it also appears to be a key feature of the poorly described biliary agenesis subgroup as well.14 The range and nature of cardiac anomalies we encountered in infants with BA appear variable with no clear pattern or aetiological predispositions. Maternal diabetes is a known association in BASM,2 and it is also a well-established association in particular cardiac anomalies such as double outlet right ventricle, truncus arteriosus and transposition of the great vessels.15 However, we could not define a clear relationship of maternal diabetes to either BASM or the cardiac anomalies in this series.16 Two genetic defects have been linked to BASM so far; variants in the polycystic kidney disease 1 like 1 (PKD1L1) and a single-nucleotide polymorphism in the Cripto, Frl-1 and Cryptic (CFC1) gene.17 18 However, neither gene has been linked with congenital cardiac anomalies and the defects were present in only a relatively small proportion of patients BASM (~2%),18 making any genetic explanation of cardiac anomalies in patients with BA unresolved.
The combination of BA and congenital cardiac defects is associated with a particularly high mortality risk and is one of the main reasons for mortality in BA overall. In our series, a third of our highest-risk groups died within 6 months, and half of infants in group B died before their first birthday. In a recent review of the UK National Cohort of Children with Serious Congenital Heart Defects, mortality in the first year of life was reported in 20%.12
We have tried to improve outcome in this high-risk group by adopting a policy of ‘Heart First’ intervention to improve overall cardiorespiratory physiology and then enable effective biliary surgery with evidence of success. Performing KPE in high-risk groups prior to cardiac surgery was associated with poor outcome and possibly delays in actually performing the cardiac intervention. By contrast, up-front cardiac surgery did not delay KPE unduly and certainly clearance of jaundice and survival were better in such infants though numbers are currently small. Early cardiac intervention in the neonatal period has been reported as safe in low birthweight infants with serious congenital cardiac anomalies,19 and the argument of minimum weight requirement prior to performing corrective or staged cardiac surgery and its effects on outcomes has been challenged in several recent articles.20–22 An analysis of The New England Regional Infant Cardiac Program showed an increased risk of mortality following cardiac surgery only in infants weighting <2 kg.23 Similarly, a recent US national analysis of 1112 infants that underwent complete tetralogy of Fallot repair demonstrated favourable outcomes when the repair was performed from as early as 31 days of age.24
Our study has the following limitations. It is a single-centre experience, although from a very large centre, with relatively a small proportion of CABA. The heterogeneity in cardiac lesions in the CABA group rendered it difficult to quantify outcomes in specific congenital cardiac defects or to obtain a matched congenital cardiac defect control. The risk groups provide a general trend towards outcome, while the individual risk should be ascertained on case-by-case basis. Furthermore, our switch in timing policy occurred relatively recently and improvements in outcome may have been due to general improvements in management of cardiac anomalies.
CABA identifies a precarious group of patients with BA that requires additional attention and planning. Our data suggest improved clearance of jaundice and survival in patients with BA with serious congenital cardiac anomalies when KPE follows corrective cardiac surgery. We therefore recognise the need for a multidisciplinary approach to planning the rapid sequence of sometimes complex cardiac surgery followed rapidly by KPE. Among these patients will still be some whose KPE will not be successful and who will proceed to liver transplantation due to decompensated cirrhosis. Within months, they will probably be unsuitable for cardiac bypass surgery. It is therefore essential that early cardiac surgery has been fully corrective to allow transplantation for those who will ultimately need it.
References
Footnotes
Contributors BA: collected data and provided the initial analysis, drafted and revised the manuscript. VG: collected data and reviewed and revised the manuscript. KP and OM: reviewed initial cardiac investigations and revised the manuscript. AB: assisted in data collection, reviewed and revised the manuscript. MD: conceptualised the study, performed data analysis, reviewed and revised the manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as online supplementary information. All data relevant to this manuscript would be available on request from the corresponding author.