Awareness, referral and age at Kasai surgery for biliary atresia in Europe: A survey of the Quality-of-Care Task Force of ESPGHAN

1 INTRODUCTION

Biliary atresia (BA) is an idiopathic congenital disease resulting in progressive fibrosclerosing obliteration of the large extrahepatic and then intrahepatic bile ducts, beginning prenatally with rapid progression after birth.¹ Although rare, affecting 1/5000 new-borns in Taiwan to 1/15,000 to 1/20,000 in Europe, Canada and the USA, BA is the most common indication for liver transplantation in childhood, and the leading hepatic cause of liver failure and death.¹ Early Kasai portoenterostomy (KPE) aims to reestablish the bile flow, and suspend or slow down the development of biliary cirrhosis. The earlier the surgery is performed, the better is the survival rate with the native liver without the need for transplantation.^2-4

BA presents with prolonged jaundice, yellow as opposed to colourless urine (a common feature to all cholestatic disorders), and pale (acholic) stools in otherwise entirely well infants. Since the early 90s, public awareness has been raised about the importance of identifying new-borns with cholestasis by 2−3 weeks of age among those with prolonged jaundice, and referring them to a specialist centre for investigation and treatment, including vitamin K injection to avoid haemorrhage.^5-7 In spite of several educational programmes, the average age at KPE did not significantly change over 20 years, and remained around 60 days of life, both in the USA^{8, 9} and Europe.^{2, 4, 10}

The joint evidence-based guidelines of European Society of Paediatric Gastroenteroloy and Nutrition (ESPGHAN) and North American Society of Pediatric Gastroenteroloy and Nutrition (NASPGHAN) from 2017 on the evaluation of cholestatic jaundice in infants, recommend the measurement of total and conjugated (direct) bilirubin in non-breastfed neonates with visible jaundice at 2 weeks of age, and no later than 3 weeks of age if breastfed.¹¹ An elevated serum conjugated bilirubin level (>1 mg/dL or >17 µmol/L) warrants timely referral to a paediatric liver specialist. The above cut off is highly sensitive to detect BA in infants between 3 and 60 days of age.¹² In practice, this recommendation is not followed. In 2020, ESPGHAN council initiated a Quality of Care (QoC) task force initiative, with the main goal to survey the implementation of diagnostic and management ESPGHAN guidelines. If deficits are identified, targeted education should further promote the recommendation in clinical practice. The Hepatology Committee of ESPGHAN selected early case finding of infants with BA as the first topic. To date neonatal dried blood spot screening is not available for this condition, although a pilot study was promising measuring matrix metalloproteinase-7 (MMP-7) levels in BA and control patients from stored dried blood spots taken within the first 3 days of life.^{13, 14} We aimed to assess the age of referral for possible KPE in a representative sample of infants and providing the rationale for referring infants for possible BA as early as possible given the improved outcome if KPE is performed before 4 weeks of age.⁴

In contrast to jaundice, an acholic stool is quite specific for obstructive cholestasis. The stool colour card (SCC picturing normal and abnormally pale or depigmented stools) (Figure 1) is noninvasive, cost effective, simple and suitable for mass screening.¹⁵ A systematic review and meta-analysis of intervention trials concluded that the SCC reduces the age at KPE to around 60 days in countries with a pre-intervention age of more than 70 days at surgery.¹³ The SCC was included in the items systematically assessed during the ‘well baby visits’ of the first weeks of life, first in Taiwan in 2002, followed by Japan, China, Canada and recently in Switzerland, France and Germany.^{13, 16-23} In Taiwan, the SCC improved the prognosis of infants with BA due to timely surgery.²¹ Screening induced anxiety in parents was not demonstrated in a Swiss survey after SCC introduction.²⁴

Here we report the results of two surveys as initial steps of the QoC Task Force activity on BA: first, to assess the age of referral and KPE in a representative sample of infants with BA in Europe; second, to survey the visibility and acceptance among French paediatricians of the recently initiated screening for BA with the SCC.

2 METHODS

3 RESULTS

3.2 Second survey: Perceived usefulness of the SCC by paediatricians

The SFP and the AFPA have about 750 and 1500 members, respectively. In total, 392 paediatricians (16%) completed the survey. Most respondents (96%) had seen the SCC in the booklet.

Half of the participants (53%) felt familiar with the target diseases and their symptoms, but 40% admitted to have little knowledge and 7% reported hardly any or no knowledge of the diseases screened with the SCC. When they were asked what diseases would be screened by the SCC, most paediatricians correctly chose cholestasis (78%) and BA (97%), while only a small fraction incorrectly chose jaundice (6%), or others (4%) (reported as free text: e.g., melena, rectal bleeding, hepatitis, cow's milk allergy, infections).

The subsequent questions explored the paediatricians' compliance with this new screening tool and how the implementation of the SCC had changed their practice (Figure 2A–E). When asked how often they inquired about the infant's stool colour, referring to the scale, during the first medical check-up visit (at age 2 weeks to 3 months), 59% ask always (at each visit), and a further 24% usually, while the remaining 17% ask the parents sometimes (9%), rarely (5%) or never (3%) (Figure 2A). In case of an abnormal stool colour, 61% would strongly agree to order a blood test, and 32% would agree, while 7% do not agree (Figure 2B). A high percentage (87%) would already ask a specialist for advice (Figure 2C). When the participants were asked whether the SCC has helped to diagnose treatable diseases in neonates, 62% considered it as very much or much helpful, while the remainder were more or less sceptical (Figure 2D). For the question of the impact of the SCC on their knowledge of neonatal jaundice, 45% reported that it improved (Figure 2E).

4 DISCUSSION

Improvement in QoC for children with neonatal cholestasis, in particular those with BA, is one of the main goals for paediatric hepatologists. The most comprehensive longitudinal data come from the French cohort covering 1428 infants diagnosed with BA between 1986 and 2015.^{4, 10} More recent data are only available from selected patients with isolated BA, excluding those with splenic malformation syndrome, comparing the incidence of BA prior and during the lockdown during the COVID pandemic.²⁵ The first step of our QoC initiative was thus to get a broad picture of case finding and age of KPE of infants with BA in Europe. While two decades ago, the goal was to perform KPE before the age of 60 days, large cohort studies indicate a highly significant inverse relationship between age at surgery and the survival rates with native liver, with the highest rate when the KPE is performed within the first 30 days of life.^{2, 4, 10}

Our European survey covered 772 infants diagnosed in the years from 2015 until 2019 and disclosed a median age at Kasai surgery of 60 days (IQR from 54 to 67 days). In addition, 6% had been referred too late to perform KPE due to advanced age and cirrhosis, which is similar to the 5.6% reported in the French cohort.⁴ Considering about 5.1 million infants born annually in the EU plus UK and Switzerland, and an incidence rate of BA in 1 of 18,000, about 1.416 infants with BA are expected during the 5-year period. Our cohort of 772 infants from all over Europe accounts for about 55% and therefore, seems quite representative for all infants with BA born in Europe during these years. Data from the French cohort 1986 and 2015 of 1340 infants disclosed median age at KPE of 59 days (range 6−199 days) with no significant improvement over 3 decades.⁴ The results of our survey confirmed the stagnation for age at surgery. Although many other factors are important (e.g., anatomic pattern, isolated vs. syndromic form, surgeon's experience, postoperative management), age at surgery is most significant for outcome and the need for liver transplantation.^{2, 26}

The median age at referral to the KPE performing centre was 53 days (IQR 48−60), with a calculated median of 7 (IQR 3−10) hospital days in the tertiary centres before surgery. The delay in the centres can be decreased by optimizing the structure, resources and process of diagnostic work up and around surgery. In contrast, the time point to identify an infant with cholestasis, further diagnostic work-up outside and final referral to the tertiary centre for surgery depends on the healthcare system of the country for example, time points for first well baby visit after discharge from maternity unit and awareness of paediatricians, general practitioners, midwifes or other healthcare professionals to initiate timely blood testing for conjugated bilirubin and speed up further work-up and referral. Measurement of conjugated bilirubin of jaundiced babies should be targeted to ‘well child visit’ if performed at 2 and not later than 3 weeks of age. In countries with later or no regular check-up, different strategies need to be established.

Several interventional trials aimed to reduce the age at KPE applying different screening strategies.¹³ One study from Texas measured conjugated bilirubin as first step in all jaundiced neonates within 60 h after birth, with retesting until 2 weeks of age if values were >95th percentile. Infants with a conjugated bilirubin of >1 mg/day or >17 μmol/L were referred for diagnostic work up.²⁷ This approach reduced the mean age at KPE from 56 days prior intervention to 36 days of age with a group difference of 19 days [95% CI: 7–32] (p = 0.004). Most intervention trials used the SCC for screening with blood sampling if stool colour is abnormal, which is less costly.¹³ A recent meta-analysis of eight such studies from Taiwan, Japan and China revealed a significant lower age at KPE in settings with a mean age at surgery of >70 days. However, in countries with a pre-screening age at KPE of <60 days, no significant reduction could be achieved. Multiple reasons for no further improvement may explain this: later development or fluctuation of abnormal stool in some infants, delayed detection and reporting by caregivers, and delays within the health care system to reach the diagnosis and referral to surgery, in part due to insufficient knowledge of healthcare professionals.

The second survey aimed to gather experience and awareness about the SCC from paediatricians practicing in France, where the card had been introduced about 2 years ago. The answers of almost 400 paediatricians to simple questions disclosed the rather low level of knowledge about cholestatic diseases in infancy, despite previous awareness campaigns in France. Reassuring was the high visibility of the SCC, its routine use, and the adequate resulting management: calling for advice. The survey has several limitations, particularly the low response rate of 16% which cannot exclude a selection bias. We also cannot exclude the results from self-assessment reflect real performance. An evaluation of the effectiveness to improve long-term outcome of infants with BA after introduction of the SCC—as paper version or digital application²⁸ as currently tested—will hopefully be reported from more countries where it has been introduced.

Besides the urgency for surgery in BA, other complications call for an early diagnosis of neonatal cholestasis regardless of the underlying disease. Malabsorption of fat-soluble vitamins, including vitamin K, impairs coagulation in spite of postnatal oral vitamin K supplementation increasing the risk of cerebral bleeding, a dreadful but preventable complication.⁵

Physiological jaundice in neonates is common, and prolonged jaundice in exclusively breastfed babies is most often not due to cholestasis, therefore, the risk of missing cholestatic diseases is very high. The knowledge of the metabolism and routes of elimination of bilirubin is poor, as the survey confirmed, and the relationship with urine or stool colour is at best confused. In the dark skinned babies, jaundice is more easily missed. Many general practitioners, paediatricians and midwifes are not aware of the current guidelines on infantile jaundice.⁹ Therefore, the QoC task force has developed a case-based awareness campaign which will be delivered to all European countries in their respective languages for distribution to health care professionals to deliver the main messages of the ESPGHAN guidelines (https://www.espghan.org/our-organisation/Quality-of-Care-Initiative).

In conclusion, our survey covering about half of all infants with BA born in the European Union, UK and Switzerland during a 5 year period disclosed a mean age at Kasai surgery of 60 days. These figures are similar to the French data obtained over the last 30 years showing no improvement over that period. Better knowledge of current guidelines and implementation of structures in the different countries which allow case finding by latest 3 weeks of age and standards in diagnostic work up and timely referral pathways, would benefit not only infants with BA, but also neonates with other diseases causing cholestasis.

THE ESPGHAN QUALITY-OF-CARE TASK FORCE

Piotr Czubkowski, Cristina Campos Goncalves, Paula Crespo Escobar, Nabil El-Lababidi, Konstantinos Gerasimidis, Katharina Ikrath, Angelika Kindermann, Sibylle Koletzko, Florence Lacaille, Thu Giang Le Thi, Anna Litwin, Emanuele Nicastro, Tena Niseteo, Rouzha Pancheva.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.