Introduction

Alagille Syndrome (ALGS) is a rare multisystem disorder with an estimated incidence rate ranging from 1 in 30,000 to 1 in 100,000 individuals [1]. It is inherited in an autosomal dominant manner in approximately 40% of patients [2]. ALGS is primarily attributed to pathogenic variants in the jagged canonical Notch ligand 1 (JAG1) and Notch Homolog 2 (NOTCH2) genes, which encode molecules involved in the NOTCH signaling pathway [3,4,5]. The severity of ALGS can range from asymptomatic presentations to fatal complications, with a mortality rate reaching up to 10% [4, 6, 7]. The clinical manifestations of ALGS are diverse, including cholestasis, characteristic facies, posterior embryotoxon, abnormal development of the heart and bones, kidney and vascular abnormalities in some patients, and liver histopathology may reveal the absence of intrahepatic bile ducts [2]. However, due to the variable expressivity of clinical manifestations among ALGS patients, some exhibit only a single typical symptom, and some are even in a subclinical state when seeking medical attention [8,9,10,11,12]. Consequently, relying on classical diagnostic criteria may lead to overlooking the diagnosis of ALGS [13, 14]. The current diagnosis primarily relies on the revised diagnostic criteria introduced in 2007 [15, 16], which incorporate genetic testing to improve diagnosis accuracy. We retrospectively analyzed the clinical data of 17 children diagnosed with ALGS at our hospital over the past 10 years, aiming to provide compelling evidence supporting the early detection of ALGS.

Materials and methods

Patients

The study enrolled 17 children diagnosed with ALGS who were admitted to the Fifth Medical Center of the PLA General Hospital from January 2012 to January 2022. ALGS diagnosis was established based on the revised criteria [15, 16], excluding children with comorbidities. Cholestasis was defined by elevated serum direct/conjugated bilirubin levels (> 1.0 mg/dL or > 17µmol/L), or by gamma-glutamyl transpeptidase (GGT) levels surpassing 3 times the upper limit of normal [17, 18]. Characteristic facies is described as an inverted triangular face, defined by a prominent forehead, deep-set eyes with hypertelorism, a straight nose with a bulbous tip, and a pointed chin [2].

Clinical data were gathered from medical records and follow-up visits. At diagnosis, demographic characteristics and clinical data were recorded, including clinical symptoms, laboratory data (liver function and lipid tests), physical examination results (growth and development condition, facial features, and indications associated with chronic hepatitis, such as jaundice, spider nevi, or palmar erythema), routine examinations (cardiac and abdominal ultrasound, chest and abdominal Computed Tomography, abdominal Magnetic Resonance Imaging, and spinal X-ray), and specialist examinations (ophthalmic slit-lamp examination). Additionally, results from liver biopsy and genetic testing were obtained. The study was ethically approved by the Ethics Committee at the Fifth Medical Center of the PLA General Hospital. All clinical information and images presented in this report were acquired with prior written consent from the patients’ parents.

Percutaneous liver pathology

Percutaneous liver biopsies were performed on 15 cases. The paraffin-embedded tissue was sectioned and stained with hematoxylin, eosin, Masson’s trichrome, cytokeratin-7 and cytokeratin-19. The histological examination was reviewed by experienced pathologists. Bile duct paucity was defined as the absence of interlobular bile ducts in at least 50% of the portal tracts within liver samples containing 10 or more portal areas [19].

Genetic variation analysis

DNA was extracted from peripheral blood, with detailed methodology outlined in the previous publications [20, 21]. Briefly, the sequenced data were then aligned with the human reference genome version 19 (hg19) using the Burrows-Wheeler Alignment (BWA) Tool (http://bio-bwa.sourceforge.net/). The Genome Analysis Toolkit (GATK) software (www.broadinstitute.org/gatk) was used to analyze single-nucleotide polymorphisms (SNPs). ANNOtation VARiants (ANNOVAR) was utilized for annotating candidate variants(annovar.openbioinformatics.org/en/latest/). MutationTaster [22], Sorting Intolerant From Tolerant (SIFT) [23, 24] and Polyphen-2 [25] programs were used to predict the potential impacts of single-nucleotide variants (SNVs). Following the American College of Medical Genetics and Genomics (ACMG) guideline [26], pathogenic variants were classified as benign, likely benign, variants of unknown clinical significance (VUS), likely pathogenic and pathogenic. The candidate causal variants identified via whole-exome sequencing (WES) were subsequently confirmed using Sanger sequencing. Co-segregation analyses were also performed within the patient’s family. Fragments covering mutated sites were amplified, purified with the Zymoclean Polymerase Chain Reaction (PCR) Purification Kit (Zymo Research, USA), and sequenced using an ABI 3730 DNA Sequencer (SeqGen, CA). Sanger sequencing data were analyzed by Chromas Lite v2.01 (Technelysium Pty Ltd., Tewantin, QLD, Australia).

Treatment and follow-up

Except for 2 patients who opted out of medication owing to their mild symptoms, the other 15 patients underwent prolonged oral therapy, which comprised of ursodeoxycholic acid, cholestyramine, bicyclic alcohol, and compound glycyrrhizin tablets, as well as various vitamin supplements. Children exhibiting severe cholestasis or pruritus underwent liver transplantation. Every three months, the children’s liver function, biochemistry, blood routine, and abdominal ultrasound were reviewed, along with regular assessments of their growth and development. Treatment adjustments were made in accordance with changes in the patients’ conditions. All patients were followed up until August 2023.

Statistical analysis

Continuous variables were shown as medians and interquartile ranges. Categorical variables were presented as counts and percentages. Statistical analysis was performed with R Foundation for Statistical Computing software (v. 4.1.3; Vienna, Austria; http://www.r-project.org/).

Results

Clinical characteristics

This study included a total of 17 participants, with 10 males and 7 females. The median age of onset among the 17 children was 2 months (range: 1-168 months), with most developing the disease within 3 months of birth. The median age at diagnosis was 5 years (range: 1–15 years).

All patients in the study exhibited liver involvement. Cholestasis was the most prevalent characteristic, affecting 94% of the patients (16/17). Characteristic facies was the second most prevalent, presenting in 88% of patients (15/17) (Fig. 1). Heart disease was identified in 75% of patients (12/16), and butterfly vertebrae were detected in 71% (12/17) (Fig. 2). The posterior embryotoxon was detected in approximately 58% of patients (7/12) assessed. Growth retardation was observed in 53% of the children (9/17), and pruritus was reported by 42% (7/17). Only Patient 8 had a relevant family history, specifically a history of ALGS in an elder brother. Abdominal MRI scans revealed cirrhosis with ascites in patients 5 and 11, which was histologically confirmed. Both patients, aged over 10 years, exhibited significant cholestasis upon admission. Detailed clinical information for each patient is provided in Table 1 and Supplementary Table 1.

Fig. 1
figure 1

Characteristic facies of Alagille syndrome in case 2: a child aged 1 year and 7 months with a prominent forehead, deep-set eyes with hypertelorism, a straight nose with a bulbous tip, and a pointed chin

Full size image
Fig. 2
figure 2

Butterfly vertebrae in case 11 (The lesions occurred in the thoracic vertebra region, as indicated by an arrow)

Full size image
Table 1 Clinical characteristics of 17 children with ALGS
Full size table

Biochemical results

Laboratory examination indicators for all patients are presented in Table 2. AST, GGT, and LAP levels were elevated in all patients. Total biliary acid (TBA) was increased in 16 patients (94%), LAP in fifteen patients (88%), alanine aminotransferase (ALT) in thirteen patients (76%), total cholesterol (TC) in thirteen patients (76%), triglycerides (TG) in twelve patients (71%), DBil in seven patients (42%), and alkaline phosphatase (ALP) levels in five patients (29%).

Table 2 Biochemical characteristics of 17 children with ALGS
Full size table

Histological results

15 children underwent liver biopsy for pathology examination, revealing intrahepatic bile duct paucity in 12 cases (Fig. 3). Mild liver inflammation and fibrosis were both present in 14 patients. The liver biopsy of one patient indicated cirrhosis (Table 3 and supplementary Table 2).

Fig. 3
figure 3

The histopathological change of liver biopsy in case 11. A: Liver cell swelling, fibrous tissue proliferation and few bile duct in the portal area (HE, original magnifi cation × 100); B: Few bile duct in the portal area (CK-7/19, original magnifi cation × 100); C: Few bile duct in the portal area (CK-7/19, original magnifi cation × 200)

Full size image
Table 3 Pathological results of liver puncture in 15 childrena with ALGS
Full size table
Table 4 Genetic traits of children with Alagille syndrome
Full size table

Genetic test results

Among the 15 individuals subjected to genetic testing, 13 carried pathogenic variants in the JAG1 gene, while none exhibited pathogenic variants in the NOTCH2 gene. No JAG1 gene pathogenic variants were detected in two patients. Thirteen distinct pathogenic variants in the JAG1 gene were identified in the 13 children harboring such variants, comprising six frameshift, three nonsense, two missense, one splice-site mutation, and one large fragment heterozygosity loss. Four novel pathogenic variants in the JAG1 gene were identified in our study. All pathogenic variants identified in our study were either previously reported or predicted to be pathogenic. The amino acid changes induced by these pathogenic variants are located at diverse positions on the JAG1 protein. Bioinformatics software was used to assess the pathogenicity of an unreported missense mutation, c.548T > A (p.I183N), which was deemed harmful by MutationTaster, SIFT, and Polyphen-2 (Table 4, Fig. 4 and Fig. S1).

Fig. 4
figure 4

The protein structure of mutation p.I183N was shown in red (PDBID:4CC0)

Full size image

Management and follow-up

After a period of treatment, most children demonstrated improved liver function test results, yet their pruritus did not experience significant relief. Liver transplantation was performed for case 8 (presenting with severe pruritus and stunted growth) and case 11 (suffering from severe cholestasis).

All patients underwent follow-up for a median duration of 7 years. During this period, 2 patients were lost to follow-up, 11 remained stable, and 4 experienced deteriorated. None of the children in the study developed liver cancer or succumbed to the disease. Apart from cases 5 and 11, who were initially diagnosed with decompensated cirrhosis, no other child progressed to this condition. Cases 8 and 11, who underwent liver transplantation, were monitored for 2 years post-surgery. Presently, both are administered tacrolimus orally as an anti-rejection measure. Although their symptoms of jaundice and pruritus resolved post-transplantation, growth and development of case 8 remained below normal levels, while case 11 exhibited postoperative anemia.

Discussion

ALGS, a multisystem disorder affecting multiple organs, was first reported by Alagille in 1969, with classical diagnostic criteria were established in 1975 [27]. The classical diagnostic criteria require histological confirmation of bile duct paucity and with a minimum of three additional characteristic symptoms [13, 14, 27]. Due to the stringency of the classical diagnostic criteria, only 12 children in our study met the criteria. Currently, the diagnosis of ALGS predominantly relies on revised criteria [15, 16] that emphasize the diagnostic significance of JAG1 gene pathogenic variants, enhancing diagnostic efficiency.

Cholestasis is widely regarded as a characteristic symptom in children with ALGS [2, 28,29,30]. In this study, 94% of patients presented with cholestasis during consultation. This percentage is comparable to the 89% incidence of cholestasis reported in the King’s College case series [31]. Bile duct paucity or hypoplasia is a key symptom of ALGS. Among the 15 patients who underwent liver pathology examination in this study, 13 (87%) patients had bile duct defects of varying degrees, a higher proportion than the 75% incidence of bile duct paucity reported by King’s College [31]. However, other studies indicate that the incidence of bile duct paucity in children with ALGS can reach up to 85% [32]. Overall, bile duct paucity is the most consistent feature of ALGS [33], and different detection rates may be related to subject age and errors in liver biopsy. To differentiate biliary atresia (BA), biliary exploration was performed on two children (cases 3 and 7) at 2 months of age. There have been reported cases where ALGS was erroneously diagnosed as biliary atresia, ultimately leading to death or liver transplantation following Kasai surgery [34, 35]. Hence, in cases of suspected biliary atresia, biliary exploration is crucial to exclude ALGS.

Characteristic facies is a well-known prominent feature of patients with ALGS, particularly those harboring pathogenic variants of the JAG1 gene [36, 37]. Characteristic facies can be identified in 80% of children with ALGS [38], while in this study, 88% of the patients exhibited this typical feature. Prompt identification of suspicious facial features in children is crucial. However, characteristic facies may not be apparent in early infancy [39, 40]. Furthermore, the use of characteristic facies in diagnosis remains controversial due to the subjectivity of facial observation and variations in observer interpretation [39]. Heart disease is the most common extrahepatic manifestation of ALGS, with an overall incidence of 94% [33]. Among these, the pulmonary artery is the most common site of anomalies, accounting for approximately 76% of all patients with cardiac anomalies [41]. In this study, approximately 75% of patients exhibited cardiac anomalies, whereas only 25% presented pulmonary artery anomalies. Although complex cardiac lesions are recognized as a significant cause of mortality in children with ALGS [32], none of the children in our follow-up study succumbed. The prevalence of butterfly vertebrae in our study was 71%, aligning with previous reports that ranged from 33–87% [42]. Most skeletal abnormalities in the studied children were predominantly found in the thoracic vertebrae, with only case 13 showing lumbar and sacral vertebrae affected. Posterior embryotoxon is the most common ocular abnormality in patients with ALGS, which usually does not affect vision and occurs in 56-95% of ALGS patients [43, 44]. Among the 12 children who underwent ophthalmic slit lamp examination in this study, 7 were found to have posterior embryotoxon, representing approximately 58% of the examined population.

Genetic testing in this study revealed a pathogenic variant detection rate of 87% for the JAG1 gene, whereas no pathogenic variant was identified in the NOTCH2 gene. These findings are consistent with the conclusions of previous studies [12]. As an important ligand of the Notch signaling pathway, changes in the structure and function of JAG1 protein significantly affect Notch signaling levels [12]. The NOTCH signaling pathway has been found to be involved in the development of multiple systems, often manifesting as multi-system symptoms in patients with ALGS [45]. However, genetic testing in this study yielded negative results for two children. This phenomenon has also been reported in previous studies [27, 46, 47], we speculate that it may be related to the fact that whole-exome sequencing technology is prone to miss deletions. Additionally, we noticed that four patients inherited their JAG1 gene pathogenic variants from their “healthy” mothers. These mothers exhibited no overt clinical symptoms during their visits, and due to limitations, liver pathology examinations were not conducted on these apparently “healthy” mothers. We speculate that the asymptomatic carriage of JAG1 gene pathogenic variants may be due to differential expression of the JAG1 gene in hepatoblasts and portal vein mesenchyme (PVM). Previous studies have suggested that the absence of Jag1 in hepatoblasts does not affect bile duct development in mice. In addition, the correlation between ALGS genotype and phenotype remains unclear, and some modifier genes are thought to impact the phenotype of patients with ALGS [48].

Ursodeoxycholic acid is currently the first-line therapy for ALGS and has been proven effective in controlling patients’ pruritus and xanthoma formation [49]. Although ursodeoxycholic acid treatment yielded positive therapeutic outcomes for these children, a subset experienced minimal relief from pruritus. To address this issue, we added choline amine as adjuvant therapy, which effectively controlled the children’s pruritus. Clinically, however, liver transplantation remains necessary for 21-31% of patients due to severe pruritus [50], with our study observing a 29% rate. While liver transplantation significantly improves patients’ original symptoms, postoperative complications and the long-term need for immunosuppressants significantly affect their quality of life, echoing findings from previous studies [51, 52]. Due to advancements in treatment modalities and improved survival rates following liver transplantation (LT), over 90% of children with ALGS now survive into adulthood [2, 29]. By the end of the follow-up, all study patients had reached adulthood with stable disease conditions. Hence, it is imperative for adult hepatologists to possess a thorough understanding of the clinical manifestations and therapeutic approaches associated with ALGS. Such knowledge is essential to address the evolving healthcare demands of ALGS patients transitioning from pediatric to adult medical care [2].

This study had several limitations. Firstly, as a retrospective, single-center study, it was constrained by a limited sample size. Secondly, the study’s subjects were exclusively hospitalized patients with relatively severe conditions, potentially introducing bias. In future studies, we plan to expand the sample size and include more outpatient cases. Lastly, all children in this study underwent whole exome sequencing for genetic testing. The limitations of this sequencing technology may result in missed detections of deletions. Furthermore, it is crucial to acknowledge the challenges posed by the homology in covering exons 1–4 of the NOTCH2 gene, which may potentially lead to their omission during exome sequencing.

Conclusion

Cholestasis is the most prevalent characteristic observed in children diagnosed with ALGS. Genetic testing should be conducted for children presenting with cholestasis, followed by the application of revised diagnostic criteria for ALGS to minimize misdiagnosis. Pharmacological therapy has proven effective for patients with ALGS, while liver transplantation may be considered in cases of severe pruritus.