UGT1A1-related disorders in Bahrain: A genetic and clinical overview of Crigler-Najjar and Gilbert syndromes

Abstract

BACKGROUND

UGT1A1-related disorders, including Crigler-Najjar syndrome (CNS) type I, type II, and Gilbert syndrome (GS), disrupt bilirubin metabolism, resulting in unconjugated hyperbilirubinemia.

AIM

To evaluate the prevalence of UGT1A1-related disorders and compare clinical features, genetic variations, management, and outcomes across the three phenotypes.

METHODS

We conducted a retrospective cohort study at the Department of Pediatrics, Salmaniya Medical Complex, Government Hospitals, Bahrain, between 2004 and 2024. This hospital is the only national referral center for genetic services in Bahrain. Of 55 suspected cases, 35 patients (63.6%) were genetically confirmed to have UGT1A1-related disorders. The three phenotypes were compared regarding clinical characteristics, biochemical parameters, genetic variations, treatment, and outcomes. Fisher’s exact test, Student’s t-test, the Mann–Whitney U test, and paired t-test were used for comparison.

RESULTS

The estimated national disease prevalence was 2.2 per 100000 individuals. GS was the most common phenotype (n = 27, 77.1%), followed by CNS-II (n = 7, 20.0%) and CNS-I (n = 1, 2.9%). Parental consanguinity was reported in 40.0% of cases. CNS-I was diagnosed earlier, at 6 months of age, than CNS-II [11.3 (0.3-16.8) years] and GS [13.4 (11.9-16.9) years]. Glucose-6-phosphate dehydrogenase deficiency was the most common coexisting condition (48.6%). Total bilirubin level at presentation was higher in CNS-I (409 μmol/L) than CNS-II (105.3 ± 46.2 μmol/L) and GS (88.2 ± 53.8 μmol/L). Homozygous (TA)7/7 polymorphism in the UGT1A1 promoter was the most common variant (91.4%). CNS-I required continuous phototherapy then liver transplantation, achieving sustained bilirubin normalization over 16.7 years of follow-up. Among patients with CNS-II, 85.7% were treated with phenobarbital, whereas 48.1% of patients with GS received phenobarbital before diagnosis, with poor adherence in 78.9% of cases. In CNS-II, there was no significant difference between indirect bilirubin level at presentation and at follow-up (92.3 ± 47.6 μmol/L vs 90.9 ± 58.0 μmol/L, P = 0.969).

CONCLUSION

Although rare, UGT1A1-related disorders represent important causes of unconjugated hyperbilirubinemia in Bahrain, with GS being the most prevalent phenotype. Genetic testing is essential for accurate diagnosis and appropriate management, particularly in patients with coexisting hematological conditions.

Key Words: Indirect hyperbilirubinemia; UGT1A1-related disorders; Crigler-Najjar syndrome; Gilbert syndrome; Bahrain

Core Tip: This first study from Bahrain demonstrated that UGT1A1-related disorders represent important causes of persistent unconjugated hyperbilirubinemia, affecting 2.2 per 100000 individuals. Gilbert syndrome (GS) was the most common phenotype, followed by Crigler-Najjar syndrome (CNS) type I and type II. Early genetic testing is crucial for diagnosis, particularly in patients with parental consanguinity and those with associated hematological disorders, mainly glucose-6-phosphate dehydrogenase deficiency. Homozygous (TA)7/7 was the most common variant. CNS-I requires lifelong phototherapy or liver transplantation, whereas CNS-II is treated with phenobarbital; however, treatment adherence remains challenging. No treatment is required for GS.

INTRODUCTION

Prolonged unconjugated hyperbilirubinemia can result from autosomal recessive inborn errors of metabolism caused by genetic defects in the UGT1A1 gene, which is located on chromosome 2[1-3]. The UGT1A1 gene encodes the enzyme uridine 5′-diphospho-glucuronosyltransferase (UDP-GT)[3]. The absence or decreased activity of the UDP-GT enzyme leads to the development of three phenotypes that represent a spectrum of the same disorder[1,2,4,5]. Accordingly, UGT1A1-related disorders are classified into three types: Crigler-Najjar syndrome (CNS) type I, type II, and Gilbert syndrome (GS)[1,4,6]. The differentiation among these phenotypes is based on clinical severity, residual enzyme activity, genetic findings, and response to medical therapy[1,4,5].

CNS-I is a rare disorder with an estimated worldwide incidence of less than 1 in 1000000 newborns[1]. It is commonly caused by nonsense mutations, frameshift mutations, splice-site mutations, and, less frequently, missense mutations in the UGT1A1 gene, leading to complete enzyme deficiency. In contrast, CNS-II is commonly caused by point mutations in the UGT1A1 gene, resulting in reduced enzyme activity to less than 10% of normal levels[1,6,7]. GS is considerably more prevalent than CNS, with a reported worldwide prevalence of 8%-10%[5,6,8,9]. It is caused by genetic variants that modulate UGT1A1 transcription, particularly those within the promoter region of the gene[6,9]. These variants lead to a 10%-35% reduction in enzymatic activity[6,9].

CNS-I is a permanent genetic condition in which unconjugated hyperbilirubinemia persists throughout life[1,10]. Elevated bilirubin levels increase the risk of bilirubin-induced neurological dysfunction (kernicterus), which can be prevented by daily phototherapy to maintain serum bilirubin within a safe range[1,3,10]. Treated individuals may survive beyond puberty; however, most eventually develop kernicterus, and the disease can be life-threatening[4]. Although bilirubin levels in CNS-I can be controlled by phototherapy or plasmapheresis, restoration of UGT1A1 activity can only be achieved through gene replacement therapy or liver transplantation (LTx) as definitive treatment[1,11]. Patients with CNS-I do not respond to enzyme-inducing agents such as phenobarbital[1].

CNS-II is considered a less severe form[8,12]. It has a later onset, typically within the first year of life[12]. Although unconjugated bilirubin levels are elevated, they usually do not exceed 20 mg/dL[13]. Unlike patients with CNS-I, those with CNS-II respond well to phenobarbital and generally have a favorable prognosis, allowing a normal quality of life[1,12,14].

Other pharmacological treatments for CNS include bilirubin-binding agents (calcium phosphate or orlistat), choleretics (ursodeoxycholic acid), and heme oxygenase inhibitors (tin protoporphyrin or zinc protoporphyrin)[12,15,16].

GS is a mild, intermittent condition in which jaundice appears during fasting, stress, or the use of certain medications[4,8,9]. The onset usually occurs after adolescence[9]. Serum unconjugated bilirubin levels fluctuate from less than 1 mg/dL to 5 mg/dL (17-85 μmol/L)[6,9]. Consequently, GS does not require dietary modification or specific medical therapy, as the prognosis is generally favorable[9].

The incidence of UGT1A1-related disorders is often higher in populations with increased rates of consanguineous marriage because of their autosomal recessive mode of inheritance[1,7,17]. This includes the Kingdom of Bahrain, where the overall rate of consanguineous marriages is 11.4%[18]. UGT1A1-related disorders have been studied in countries neighboring Bahrain and worldwide[1-3,19,20]. However, they have not been investigated in Bahrain. Accordingly, this study aimed to assess the prevalence and the clinical and molecular spectrum of UGT1A1-related disorders, with a focus on genotype–phenotype correlations, associated hemolytic comorbidities, management strategies, and clinical outcomes across the disease spectrum.

MATERIALS AND METHODS

Study design and setting

This retrospective cohort study was conducted at the Department of Pediatrics, Salmaniya Medical Complex (SMC), Government Hospitals, Manama, Bahrain, and included a review of medical records from March 24, 2004, to December 31, 2024. SMC is the largest tertiary healthcare facility in the country and serves as the sole national referral center for genetic services. Moreover, SMC is the only referral center in Bahrain for genetic testing in suspected cases of UGT1A1-related disorders (CNS and GS). Patients with persistent unconjugated hyperbilirubinemia received medical care either as inpatients or through outpatient gastroenterology and genetic clinics.

Study population

All patients with persistent hyperbilirubinemia who were genetically confirmed to have CNS-I, CNS-II, or GS were included in the study. Patients with persistent jaundice who did not undergo genetic testing, those whose test results were negative for UGT1A1-related disorders, and those who lacked biochemical results for bilirubin monitoring were excluded. The patient selection process is illustrated in Figure 1. Of the 55 patients with persistent hyperbilirubinemia, 35 were genetically confirmed to have UGT1A1-related disorders. The genetic diagnosis was established by targeted sequencing of the UGT1A1 gene, performed at external laboratories (Centro de Genética Clínica Genetics Unilabs, Portugal, or the Scottish Molecular Genetic Consortium, Human Genetics Unit, Department of Pathology, Level 6, Ninewells Hospital and Medical School, Dundee, United Kingdom).

Figure 1 Flowchart of genetic testing results of patients with clinically suspected UGT1A1-related disorders. CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome.

Molecular genetic analysis

After genomic DNA was extracted from peripheral blood samples, multiplex polymerase chain reaction (PCR) was performed to target the promoter region, including the A(TA)nTAA variant, as well as the coding regions and intron-exon boundaries (± 8 bp) of the UGT1A1 gene. During library preparation, tagmentation (QXT, Agilent Technologies, CA, United States) was used to simultaneously fragment the DNA samples, followed by next-generation sequencing (MiSeq, Illumina, San Diego, United States), with variant analysis performed using the DRAGEN workflow (Illumina, San Diego, United States). When necessary, Sanger sequencing was performed for regions with coverage below 15 ×. All variants were confirmed by independent PCR amplification and sequencing.

Data collection

Data were retrieved from patients’ medical records, including archived paper-based files and the hospital’s electronic database (I-SEHA). Demographic variables, including sex, nationality, age at diagnosis, age at the time of the study, gestational age, birth weight, family history of jaundice or UGT1A1-related disorders, and parental consanguinity, were collected. Symptoms such as jaundice, changes in urine and stool color, pruritus, and cosmetic concerns related to yellowish skin discoloration were documented, as were physical findings including hepatomegaly and/or splenomegaly. Associated conditions, particularly hematological disorders such as glucose-6-phosphate dehydrogenase (G6PD) deficiency, sickle cell disease (SCD), and thalassemia, were recorded. Because sequencing specifically targeted UGT1A1, other causes of unconjugated hyperbilirubinemia were evaluated through biochemical and hematological investigations. A quantitative G6PD enzyme assay was performed in all patients, and hemoglobin fractionation was analyzed using high-performance liquid chromatography. Genetic testing for hemoglobinopathies, particularly β-thalassemia, was performed in selected patients based on clinical suspicion and abnormal hematological indices. Biochemical test results at presentation and at the last follow-up visit-including complete blood count, reticulocyte percentage, liver function tests [total protein, serum albumin, total bilirubin, direct bilirubin, indirect bilirubin, alkaline phosphatase, alanine aminotransferase (ALT), and gamma-glutamyl transferase], and 25-hydroxyvitamin D levels-were reviewed. Causative genetic variants associated with UGT1A1-related disorders were retrieved from archived genetic reports. Abdominal ultrasound findings were also documented, as ultrasonography constitutes a routine component of care for patients with inherited hematological disorders at our institution. The management plan for each patient-including phototherapy; medications such as phenobarbital and ursodeoxycholic acid; blood transfusions; exchange transfusions; and surgical interventions such as cholecystectomy, splenectomy, and LTx-was assessed. Treatment response, determined by trends in serum indirect bilirubin levels, follow-up duration, and clinical outcomes, was recorded.

Statistical analysis

Patients’ data were entered into a Microsoft Excel 2016 spreadsheet and subsequently transferred to the SPSS [IBM Corp. (2012). IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY, United States] for further analysis. Categorical variables were presented as n (%). Continuous variables were expressed as means ± SD for normally distributed data or as medians and interquartile ranges (IQRs) for nonnormally distributed data. The three types of UGT1A1-related disorders were compared with respect to demographic characteristics, clinical presentations, associated conditions, biochemical and genetic findings, medical and surgical therapies, and outcomes. The CNS-I group included only one patient. Accordingly, comparative analyses were conducted between the CNS-II and GS groups, and the single patient with CNS-I was excluded from these analyses. Fisher’s exact test was used to compare categorical variables, whereas the Mann-Whitney U test and Student’s t-test were used to compare continuous variables. The Mann-Whitney U test was also applied to compare indirect bilirubin levels at the last follow-up visit between patients who received phenobarbital therapy and those who did not. In patients with CNS-II, serum indirect bilirubin levels at initial presentation and at the last follow-up visit were compared using paired t-tests. The CI was set at 95%, and a P value < 0.05 was considered statistically significant.

Ethical approval

This study was conducted in accordance with the principles of the Declaration of Helsinki (1964), as revised in 2000, and was approved by the Research and Research Ethics Committee, Government Hospitals, Manama, Bahrain (No. 42080523). As this was a retrospective study, the requirement for informed consent was waived.

RESULTS

Prevalence of UGT1A1-related disorders in Bahrain

From March 2004 to December 2024, 55 patients were suspected of having UGT1A1-related disorders. Of these, 43 (78.2%) underwent genetic testing, and 35 patients (63.6%) from 34 unrelated families were genetically confirmed to have UGT1A1-related disorders and were included in the study. CNS-I was identified in one patient (2.9%), CNS-II in seven patients (20.0%), and GS in 27 patients (77.1%). Based on Bahrain’s population statistics for June 2024, the total population was 1588670[21]. Accordingly, the overall estimated prevalence of UGT1A1-related disorders in Bahrain was 2.2 cases per 100000 individuals (0.002%), corresponding to 0.06 cases for CNS-I, 0.4 cases for CNS-II, and 1.7 cases for GS per 100000 individuals. Most cases were identified during the last five years (2020-2024) (Figure 2). The average annual incidence of UGT1A1-related disorders was 4.022 cases per 100000 live births, including 0.128 for CNS-I, 0.792 for CNS-II, and 3.101 for GS per 100000 live births.

Figure 2 Distribution of patients with UGT1A1-related disorders by year of diagnosis, 2004-2024. CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome.

Characteristics of the study cohort

Most patients were male (n = 27, 77.1%), with a male-to-female ratio of approximately 3.4:1 (Table 1). The majority were Bahraini nationals (n = 32, 91.4%), whereas the remaining three patients (8.6%) were non-Bahraini nationals (one each from Ireland, Pakistan, and Jordan). The median age at diagnosis was 13.1 years (IQR: 9.0-16.8). The single patient with CNS-I presented during the neonatal period, and the diagnosis was genetically confirmed at the age of 6 months and 24 days. The median age at diagnosis was 11.3 years (IQR: 0.3-16.8) for patients with CNS-II and 13.4 years (IQR: 11.9-16.9) for those with GS.

Table 1 Clinical characteristics of patients with UGT1A1-related disorders, n (%)/medians (interquartile range).

Clinical characteristics	Total, n = 35 (100)	CNS-I, n = 1 (2.9)	CNS-II, n = 7 (20.0)	GS, n = 27 (77.1)	P value1
Demographics
Sex					1.0002
Male	27 (77.1)	1 (100)	6 (85.7)	20 (74.1)
Female	8 (22.9)	0 (0.0)	1 (14.3)	7 (25.9)
Nationality					0.5112
Bahraini	32 (91.4)	1 (100)	6 (85.7)	25 (92.6)
Non-Bahraini	3 (8.6)	0 (0.0)	1 (14.3)	2 (7.4)
Age at the time of diagnosis (years)	13.1 (9.0-16.8)	0.58	11.3 (0.3-16.8)	13.4 (11.9-16.9)	0.1103
Age at the time of study (years)	15.2 (13.3-19.0)	21.4	15.3 (14.2-19.0)	14.9 (13.1-18.5)	0.7823
Gestational age (n = 30)					1.0002
Term	26 (86.7)	1 (100)	6 (85.7)	19/22 (86.4)
Preterm	4 (13.3)	0 (0.0)	1 (14.3)	3/22 (13.6)
Birth weight (n = 27)	3 (2.5-3.5)	3.0	3.0 (2.5-3.5)	3.0 (2.5-3.5)	0.9543
Family history of prolonged jaundice	9 (25.7)	0 (0.0)	4 (57.1)	5 (18.5)	0.0612
Parental consanguinity	14 (40.0)	1 (100)	4 (57.1)	9 (33.3)	0.3872
Clinical presentations
Jaundice	35 (100)	1 (100)	7 (100)	27 (100)	N/A
Changes in urine color	13 (37.1)	0 (0.0)	2 (28.6)	11 (40.7)	0.6822
Pruritus	10 (28.6)	0 (0.0)	3 (42.9)	7 (25.9)	0.3942
Changes in stool color	5 (14.3)	0 (0.0)	1 (14.3)	4 (14.8)	1.0002
Splenomegaly	13 (37.1)	0 (0.0)	2 (28.6)	11 (40.7)	0.6822
Hepatomegaly	7 (20.0)	0 (0.0)	1 (14.3)	6 (22.2)	1.0002
Associated diseases
G6PD deficiency	17 (48.6)	1 (100)	4 (57.1)	12 (44.4)	0.6812
Sickle cell disease	14 (40.0)	0 (0.0)	2 (28.6)	12 (44.4)	0.6722
Alpha-thalassemia	5 (14.3)	1 (100)	1 (14.3)	3 (11.1)	1.0002
Sickle-thalassemia	3 (8.6)	0 (0.0)	0 (0.0)	3 (11.1)	1.0002
Sickle cell trait	3 (8.6)	0 (0.0)	1 (14.3)	2 (7.4)	0.5112
Beta-thalassemia major	2 (5.7)	0 (0.0)	0 (0.0)	2 (7.4)	1.0002
Beta-thalassemia trait	1 (2.9)	0 (0.0)	0 (0.0)	1 (3.7)	1.0002
Eczema	10 (28.6)	0 (0.0)	3 (42.9)	7 (25.9)	0.3942

¹After excluding the single patient with Crigler-Najjar syndrome-I, the P values represent the differences between Crigler-Najjar syndrome-II and Gilbert syndrome groups only. Boldface indicates statistically significant difference with P < 0.05.

²Fisher’s exact test.

³Mann-Whitney U test.

CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome; N/A: Not applicable; G6PD: Glucose-6-phosphate dehydrogenase.

At the time of the study, 24 (68.6%) patients were children (< 18 years) [19 (54.3%) with GS and five (14.3%) with CNS-II], whereas the remaining 11 (31.4%) were adults [eight (22.9%) with GS, two (5.7%) with CNS-II, and one (2.9%) with CNS-I]. A family history of prolonged jaundice was negative in the patient with CNS-I but positive in four patients (57.1%) with CNS-II (mother, brother, sister, and uncle; n = 1 each). Among patients with GS, five (18.5%) had at least one affected family member (two sisters and one each of father, brother, uncle, and cousin). One relative of a patient with CNS-II (a brother) was genetically confirmed to have CNS-II and was included in the study. The remaining family members had not undergone genetic testing at the time of the study. Parental consanguinity was reported in 40.0% of cases (n = 14).

Clinical presentations and associated diseases

All patients presented with yellowish discoloration of the skin and sclera; however, the severity of jaundice varied among individuals, and some patients had persistent jaundice that significantly affected their social lives. Jaundice-related cosmetic concerns were reported in 16 patients (45.7%), either by the patients themselves (n = 10, 28.6%), by family members (n = 4, 11.4%), or by both (n = 2, 5.7%). Splenomegaly and hepatomegaly were observed mainly in patients with GS (40.7% and 22.2%, respectively). G6PD deficiency was the most frequently associated condition (n = 17, 48.6%).

Biochemical, genetic, and radiological findings

At presentation, all patients had low hemoglobin levels and elevated reticulocyte percentages (Table 2). The mean serum protein level was within the normal range in all groups; however, it was lower in the patient with CNS-I than in those with CNS-II or GS (62.0 g/L vs 70.3 ± 3.9 g/L vs 74.8 ± 5.9 g/L, respectively). After excluding the single patient with CNS-I, the difference between the CNS-II and GS groups was not statistically significant (P = 0.086, 95%CI: -9.1 to 0.62). Patients with CNS-I had higher total bilirubin levels than those with CNS-II and GS (409 μmol/L vs 105.3 ± 46.2 μmol/L vs 88.2 ± 53.8 μmol/L, respectively). The ALT level was elevated only in the patient with CNS-I. Vitamin D levels were low in 25 of the 27 tested patients (92.6%) [CNS-I (1/1, 100%), CNS-II (3/3, 100%), and GS (21/23, 91.3%)].

Table 2 Biochemical parameters of patients with UGT1A1-related disorders, n (%)/median (interquartile range).

Biochemical parameters	Normal range	Total, n = 35 (100)	CNS-I, n = 1 (2.9)	CNS-II, n = 7 (20.0)	GS, n = 27 (77.1)	P value1 (95%CI)
Hemoglobin (g/dL)	12.0-14.5	10.9 ± 2.1	10.9	11.0 ± 1.7	10.9 ± 2.2	0.955 (-1.8, 1.9)2
Hematocrit (%)	33-45	33.2 ± 5.8	32.1	33.9 ± 3.8	33.1 ± 6.4	0.721 (-4.3, 6.1)2
RBC (× 1012/L)	3.9-5.2	4.4 ± 0.9	5.3	4.5 ± 0.9	4.4 ± 0.9	0.732 (-0.7, 0.9)2
Reticulocyte (%) (n = 32)	0.5-1.5	2.7 (1.9-5.6)	1.9	1.8 (1.1-6.1) (n = 6)	3.0 (2.1-5.3) (n = 25)	0.3683
Platelet count (× 109/L)	150-400	286 (221-396)	263	331 (259-403)	282 (221-381)	0.3383
Total protein (g/L)	57-82	74 ± 5.9	62.0	70.3 ± 3.9	74.8 ± 5.9	0.086 (-9.1, 0.62)2
Serum albumin (g/L)	38-54	44.5 ± 3.4	37.0	44.1 ± 5.3	44.7 ± 2.9	0.706 (-3.6, 2.4)2
Total bilirubin at presentation (μmol/L)	5-21	89.7 ± 52.7	409	105.3 ± 46.2	88.2 ± 53.8	0.447 (-28.2, 62.5)2
Direct bilirubin (μmol/L)	≤ 5.0	14 (10-18)	4.0	15 (8-16)	14 (11-19)	0.4683
Indirect bilirubin at presentation (μmol/L)	≤ 18	72 (24-126)	405	95 (40-122)	68 (23-126)	0.1943
Maximum total bilirubin (μmol/L)	5-21	147 (98-228)	409	188 (121-227)	133 (91-238)	0.4063
Maximum indirect bilirubin (μmol/L)	≤ 18	122.7 (82-210)	405	173 (109-210)	119 (67-182)	0.2973
ALP (U/L)	142-335	185 (132-254)	246	254 (132-343)	180 (118-204)	0.3833
ALT (U/L)	≤ 33	17 (12-25)	106	17 (11-25)	17 (12-22)	0.9833
GGT (U/L)	≤ 38	11 (8-20)	20	15 (7-38)	10 (9-16)	0.9323
Vitamin D level (nmol/L) (n = 27)	> 50	31.3 ± 11.5	48	31.7 ± 6.0 (n = 3)	30.5 ± 11.8 (n = 23)	0.871 (-13.3, 15.6)2

¹After excluding the single patient with Crigler-Najjar syndrome-I, the P values represent the differences between Crigler-Najjar syndrome-II and Gilbert syndrome groups only. Boldface indicates statistically significant difference with P < 0.05.

²Student’s t-test.

³Mann-Whitney U test.

CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome; RBC: Red blood cells; ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; GGT: Gamma-glutamyl transferase.

Table 3 summarizes the results of genetic testing. Apart from the UGT1A1 (NM_000463.3):c.161G>A (p.Gly54Val) variant, all other variants had been previously reported[2,7,22-25]. Genetic testing of the patient with CNS-I revealed a homozygous (TA)7/7 polymorphism in the promoter region, along with a homozygous missense pathogenic variant, UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg), in exon 3. Genetic screening of the parents revealed that both were heterozygous carriers of the same UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) variant; however, their promoter region status was not documented. All seven patients with CNS-II exhibited the (TA)7/7 polymorphism. Of these, six (85.7%) were homozygous, and one (14.3%) was heterozygous, in whom this finding was observed alongside additional genetic variants. Among the 27 patients with GS, 25 (88.8%) were homozygous for the (TA)7/7 promoter variant; one (3.7%) carried an additional homozygous UGT1A1 (NM_000463.3):c.-3275T>G (p.?) variant; and two (7.4%) were compound heterozygous for the (TA)7/8 promoter polymorphism.

Table 3 Genetic variants of patients with UGT1A1-related disorders, n (%).

Polymorphism	Zygosity	UGT1A1 variant	Location	Zygosity	Classification	Total, n = 35 (100)
CNS-I						1 (2.9)
(TA)7/7	Homozygous	c.1070A>G p.(Gln357Arg)[2,7,22]	Exon 3	Homozygous	Pathogenic	1 (2.9)
CNS-II						7 (20.0)
(TA)7/7	Homozygous	c.907G>A p.(Val303Met)[24]	Exon 2	Homozygous	VUS	2 (5.7)
(TA)7/7	Heterozygous	c.161G>A p.(Gly54Val)	Exon 1	Heterozygous	VUS	1 (2.9)
(TA)7/7	Homozygous	c.674T>G p.(Val225Gly)[23]	Exon 1	Homozygous	LP	1 (2.9)
		c.907G>A p.(Val303Met)[24]		Homozygous	VUS	1 (2.9)
(TA)7/7	Homozygous	c.674T>G p.(Val225Gly)[23]	Exon 1	Homozygous	LP	1 (2.9)
		c.907G>A p.(Val303Met)[24]		Heterozygous	VUS	1 (2.9)
(TA)7/7	Homozygous	c.674T>G p.(Val225Gly)[23]	Exon 1	Heterozygous	LP	1 (2.9)
		c.907G>A p.(Val303Met)[24]		Heterozygous	VUS	1 (2.9)
(TA)7/7	Homozygous	c.1070A>G p.(Gln357Arg)[2,7,22]	Exon 3	Heterozygous	Pathogenic	1 (2.9)
GS						27 (77.1)
(TA)7/7	Homozygous					24 (68.6)
(TA)7/8	Compound heterozygous					2 (5.7)
(TA)7/7	Homozygous	c.-3275T>G[25]		Homozygous	VUS	1 (2.9)

CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome; LP: Likely pathogenic; VUS: Variant of unknown significance.

Abdominal ultrasound was performed in 26 patients (74.3%) before the genetic diagnosis of UGT1A1-related disorders; results were unavailable for three patients (11.3%). Notably, 20 of the 26 patients (76.9%) who underwent abdominal ultrasound had at least one associated inherited hematological disorder, including SCD (n = 13, 65.0%), reduced G6PD activity (n = 10, 50.0%), sickle-thalassemia (n = 3, 15.0%), alpha-thalassemia (n = 2, 10.0%), beta-thalassemia major (n = 2, 10.0%), and beta-thalassemia trait (n = 1, 5.0%). Among the 23 patients with available ultrasound results, 20 (86.9%) had abnormal findings, including splenomegaly (n = 11, 47.8%), hepatomegaly (n = 11, 47.8%), SCD-related nephropathy (n = 8, 34.8%), biliary sludge (n = 2, 8.7%), and fatty liver, prominent intrahepatic biliary ducts suggestive of biliary cirrhosis, dilated common bile duct, cholelithiasis, auto-splenectomy, hyperechoic splenic lesions, bilateral renal cysts, and a large left adrenal myelolipoma (n = 1, 4.4% each). Three patients (13.0%) had normal ultrasound findings. The patient with CNS-I had no hepatomegaly or splenomegaly. There were no statistically significant differences between CNS-II and GS in the presence of hepatomegaly [1/3 (33.3%) vs 10/20 (50.0%), respectively; P = 1.000] or splenomegaly [0/3 (0.0%) vs 11/20 (55.0%), respectively; P = 0.217].

Therapeutic approaches and outcomes

Patient management is summarized in Table 4. Initially, before diagnostic confirmation, the patient with CNS-I received phototherapy in the neonatal intensive care unit in addition to phenobarbital therapy. No exchange transfusion was required during hospitalization. After discharge, the patient received home phototherapy for nearly 10 hours daily, and phenobarbital was discontinued once the genetic diagnosis was established. At the age of 3.7 years, he was referred overseas for auxiliary living-related LTx in Saudi Arabia, with his mother serving as the donor. Post-transplantation, bilirubin levels normalized rapidly, and phototherapy was discontinued. He was followed regularly in outpatient clinics and remained clinically stable on tacrolimus-based immunosuppressive therapy. At 16.7 years after LTx, his indirect bilirubin level was 17 μmol/L. Among the seven patients with CNS-II, six (85.7%) were prescribed phenobarbital therapy; five (83.3%) demonstrated poor adherence, whereas one (16.7%) demonstrated good adherence. The reasons for poor adherence included patient refusal (n = 2/5, 40.0%), parental reluctance to administer the medication (n = 2/5, 40.0%), and discontinuation following consultation with adult gastroenterologists (n = 1/5, 20.0%). Before diagnosis and while awaiting genetic test results, 13 of the 27 patients with GS (48.1%) also received phenobarbital; adherence was poor in 10 (76.9%) and good in three (23.1%). Poor adherence was attributed to adverse drug effects (n = 4/10, 40.0%), such as lethargy and decreased concentration; discontinuation following consultation with adult gastroenterologists (n = 2/10, 20.0%); patient refusal (n = 2/10, 20.0%); symptom improvement (n = 1/10, 10.0%); and forgetfulness (n = 1/10, 10.0%).

Table 4 Management of patients with UGT1A1-related disorders, n (%).

Patient management	Total, n = 35 (100)	CNS-I, n = 1 (2.9)	CNS-II, n = 7 (20.0)	GS, n = 27 (77.1)
Medications
Phenobarbital	20 (57.1)	1 (100)	6 (85.7)	13 (48.1)
Folic acid	20 (57.1)	1 (100)	3 (42.9)	16 (59.3)
Vitamin D supplementation	20 (57.1)	1 (100)	1 (14.3)	18 (66.7)
Hydroxyurea	16 (45.7)	0 (0.0)	2 (28.6)	14 (51.9)
Ursodeoxycholic acid	13 (37.1)	0 (0.0)	1 (14.3)	12 (44.4)
Other medications1	16 (45.7)	1 (100)	1 (14.3)	14 (51.9)
Medical/surgical procedures
Blood transfusion	15 (42.9)	1 (100)	2 (28.6)	12 (44.4)
Phototherapy	13/31 (41.9)	1 (100)	4/6 (66.7)	8/24 (33.3)
Cholecystectomy	8 (22.9)	0 (0.0)	2 (28.6)	6 (22.2)
Splenectomy	6 (17.1)	0 (0.0)	1 (14.3)	5 (18.5)
Liver transplantation	1 (2.9)	1 (100)	0 (0.0)	0 (0.0)

¹Crigler-Najjar syndrome-I: Tacrolimus, magnesium supplement, and iron supplement (n = 1); Crigler-Najjar syndrome-II: Prophylactic antibiotics (n = 1); Gilbert syndrome: Prophylactic antibiotics and aspirin (n = 5, each), iron chelation (n = 4), laxative (n = 4), calcium supplement (n = 3); omeprazole (n = 2), rifampicin, erythromycin, multivitamins, fat-soluble vitamins, pancreatic enzyme replacement therapy (Creon), L-thyroxin, growth hormone, testosterone, warfarin, antihypertensives, frusemide, inhaled steroids, and bronchodilators (n = 1, each).

CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome.

The follow-up duration was available for 33 of 35 patients (94.3%) and was longer in CNS-I (16.7 years) than in CNS-II [3.2 years (IQR: 2.5-13.5)] and GS [1.2 years (IQR: 0.4-3.6)]. Patients with CNS-II had a significantly longer follow-up duration than those with GS (P = 0.021). Regardless of the UGT1A1 phenotype, the median indirect bilirubin level at the last follow-up visit was higher among patients who received phenobarbital than among those who did not [40.4 μmol/L (IQR: 20.3-87.0) vs 37.0 μmol/L (IQR: 26.9-87.0)]; however, this difference was not statistically significant (P = 0.934). In patients with CNS-II, there was no statistically significant difference between indirect bilirubin levels at presentation and at follow-up (92.3 ± 47.6 μmol/L vs 90.9 ± 58.0 μmol/L, P = 0.969).

DISCUSSION

This study demonstrated that genetically confirmed UGT1A1-related disorders are rare in Bahrain, with an estimated prevalence of 22 cases per million individuals (0.63 for CNS-I, 4.4 for CNS-II, and 17 for GS) and an average annual incidence of 40.22 cases per million live births (1.28 for CNS-I, 7.92 for CNS-II, and 31.01 for GS). CNS-I is extremely rare worldwide, with an estimated incidence of approximately 1 per 1 million live births[26]. To date, the highest reported incidence of CNS-I is in Tunisia (8 per million newborns), followed by Croatia (6 per million live births)[27,28]. The exact incidence of CNS-II remains unclear and requires further investigation. GS was the most prevalent UGT1A1-related disorder in this Bahraini cohort (0.0169%); however, this prevalence is lower than the reported global prevalence of GS, which ranges from 3% to 8.6%[9]. This discrepancy may be attributed to the higher prevalence of GS among White populations (approximately 10%) or to underdiagnosis[5].

The present study demonstrated that 88.6% of UGT1A1-related disorder cases were diagnosed during the last five years (2020-2024). However, this marked increase in the number of diagnosed patients is likely attributable to improved access to genetic testing and increased awareness among healthcare professionals regarding genetic disorders rather than to a true rise in incidence[18,29]. In Tunisia, Aronson et al[27] attributed the elevated incidence of CNS-I to high rates of consanguineous marriage (25%-60%) and to a founder effect. A founder effect was also reported by Koshy et al[2] in two children with CNS-I from Kuwait. Notably, these authors described the same molecular configuration observed in our patient with CNS-I, as both children carried the UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) mutation located in exon 3. This mutation has been detected only in Arab populations and has consistently been associated with the (TA)7 promoter polymorphism[2]. These findings strongly support the association of CNS-I with a founder effect not only in Tunisia but also in other Arab countries. Haplotype analysis conducted by Petit et al[22] suggested that the UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) mutation emerged approximately 32 generations ago in the Tunisian population and that its spread was largely confined to consanguineous marriages. Bahrain also has a relatively high consanguinity rate (11.4%)[18], and the present study identified an even higher rate among patients with UGT1A1-related disorders (40.0%). Similarly, in Saudi Arabia, a neighboring country, high rates of consanguineous marriage have been reported[1]. Although consanguineous marriage is uncommon in Croatia, isolated populations on islands and in remote villages have experienced reproductive isolation, consanguinity, and inbreeding[28].

Most patients in this study had GS (77.1%), followed by CNS-II (20.0%) and CNS-I (2.9%). However, none of the reviewed studies included all three phenotypes. Wu et al[8] from China reported 310 patients with isolated unconjugated hyperbilirubinemia; 232 (74.8%) had GS, and 78 (25.2%) had CNS-II. Similarly, in another study from China, Sun et al[30] evaluated 95 patients with unconjugated hyperbilirubinemia, of whom 59 (62.1%) were diagnosed with GS, 36 (37.9%) with CNS-II, and none with CNS-I. Five other studies included both types of CNS[14,15,20,23], whereas most previous studies examined only one phenotype, either CNS-I alone[1-3,7,16,22,28], CNS-II alone[4,13,31-33], or GS alone[9,19,34]. Tables 5 and 6 summarize the main clinical, biochemical, and genetic features of UGT1A1-related disorders reported in studies from neighboring countries and worldwide.

Table 5 Summary of the main features of UGT1A1-related disorders.

Main features	CNS-I	CNS-II	GS
Prevalence	Ultra-rare[26]	Unclear (no record)	Prevalent (8%-10%)[5,6,8,9]
UGT1A1 enzyme activity	0%[5]	< 10%[6]	10%-30%[9]
Age at presentation	Neonatal period[1,28]	Infancy or childhood[33]	Adolescence or adulthood[8]
Triggers of jaundice	None	Fasting, illness or stress[4,6]	Fasting, illness, or stress[5,8]
Clinical severity; unconjugated bilirubin range (µmol/L)	Severe[28,33]; 340-850[33]	Moderate[33]; 85-340[33]	Mild[8,33]; < 85[33]
Kernicterus risk	High[1,3,6,10]	Low[7]	None[5]
Medical therapy	Phototherapy[1,26], plasma phoresies[1,26], or liver transplantation[3,11]	Phenobarbital[33]	No treatment needed[5]
Outcome	Poor if not treated[1,3,10]	Good[33]	Excellent[6]

CNS: Crigler-Najjar syndrome; GS: Gilbert syndrome.

Table 6 Summary of UGT1A1-related disorders studies published from neighboring countries to Bahrain and worldwide.

Country	Ref.	n	Sex (M:F)	Promotor variant	UGT1A1 mutation	Zygosity	Location	Additional variant	Zygosity
CNS-I
Saudi Arabia	Nazer et al[1], 1998	12	8:4	NR	NR	NR	NR	NR	NR
Kuwait	Koshy et al[2], 2004	2	0:2	(TA)7/7	1070A>G (p.Gln357Arg)	+/-	Exon 3	NR	NR
Turkey	Ozçay et al[3], 2009	4	NR	NR	NR	NR	NR	NR	NR
Tunisia	Abdellaoui et al[7], 2022	12/17	NR	(TA)7/7	c.1070A>G (p.Gln357Arg)	+/+	Exon 3	NR	NR
		5/17	NR	(TA)8/8	c.1070A>G (p.Gln357Arg)	+/+	Exon 3	NR	NR
Tunisia	Petit et al[22], 2008	23	6:17	(TA)7/7	c.1070A>G (p.Gln357Arg)	+/+	Exon 3	NR	NR
Germany	Schröder et al[14], 2021	12/13	10:3	NR	NR	NR	NR	NR	NR
Netherlands	Hafkamp et al[15], 2007	7/16	5:11	NR	NR	NR	NR	NR	NR
China	Li et al[4], 2015	6	6:2	(TA)7/7	c.221G>A (p.Gly71Arg)	+/+	Exon 1	NR	NR
Croatia	Kovačić Perica et al[28], 2023	2/8	4:4	(TA)7/7	c.717_718delAG (p.Q239fs*256)	NR	Exon 1	NR	NR
		1/8		(TA)7/7	c.722_723delAG (p.Glu241Glyfs*16)	NR	Exon 1	NR	NR
		2/8		(TA)6/6	c.1021C>T (p.Arg341*)	NR	Exon 3	NR	NR
		1/8		N/A	c.1021C>T (p.Arg341*)	NR	Exon 3	NR	NR
		1/8		Not performed	Confirmed by chromatographic bile analysis	NR	NR	NR	NR
		1/8		Lost result	NR	NR	NR	NR	NR
Iran	Mohammadi Asl et al[20], 2013	4/12	2:2	NR	c.238_240delAG	+/+	Exon 1	NR	NR
					c.479T>A (p.Val160Glu)	+/+	Exon 1	NR	NR
Iran	Maruo et al[23], 2015	2	1:1	NR	c.381insGG (p.C127Wfs*23)	+/+	Exon 1	NR	NR
United States	Strauss et al[16], 2006	17/20	NR	NR	c.222C>A (p.Tyr74Ter)	+/+	Exon 1	NR	NR
		1/20			c.1069C>T (p.Gln357Ter)	+/+	Exon 3	NR	NR
		1/20			c.1305-1G>A, c.877_890delinsT	+/-, +/-	Exon 4, Exon 2	NR	NR
CNS-II
Japan	Maruo et al[31], 2006	1	1:0	(TA)7/7	c.115C>G (p.His39Asp)	+/+	Exon 1	NR	NR
Iran	Maruo et al[23], 2015	1	1:0	(TA)7/7	c.381insGG (p.C127Wfs*23)	+/-	Exon 1	c.674T>G (p.Val225Gly)	+/- (Exon 1)
France	Labrune et al[32], 2002	1	0:1	(TA)8/8	c.1213A>G (p.Asn400Asp)	+/+	Exon 4	NR	NR
Germany	Schröder et al[14], 2021	1/13	10:3	NR	NR	NR	NR	NR	NR
Netherlands	Hafkamp et al[15], 2007	9/16	5:11	NR	NR	NR	NR	NR	NR
China	Li et al[4], 2015	1/11	8:3	NR	c.211G>A (p.Gly71Arg)	+/+	NR	c.1456T>G (p.Tyr486Asp)	+/+
		1/11			c.211G>A (p.Gly71Arg)	+/-	NR	c.1456T>G (p.Tyr486Asp)	+/+
		1/11			c.1091C>T (p.Pro364 Leu)	+/+	NR	c.-40_-39insTA	+/-
		1/11			c.1456T>G (p.Tyr486Asp)	+/-	NR	c.1253del (p.Met418Argfs*6)	+/-
		2/11			c.715C>T (p.Gln239*)	+/-	NR	c.-40_-39insTA	+/+
		1/11			c.1091C>T (p.Pro364 Leu)	+/-	NR	c.211G>A (p.Gly71Arg)	+/-
		1/11			c.211G>A (p.Gly71Arg)	+/-	NR	c.-40_-39insTA	+/-
							NR	c.-3279T>G	+/-
		2/11			c.211G>A (p.Gly71Arg)	+/-	NR	c.-3345delC	+/-
		1/11			c.211G>A (p.Gly71Arg)	+/-	NR	NR	NR
China	Wu et al[8], 2024	78/310	50:28	(TA)7/7 (TA)7/6 (TA)6/6	c.211G>A (p.Gly71Arg)	+/+, +/-	Exon 1	NR	NR
Japan	Yamamoto et al[13], 1998	5/7	3:4	(TA)7/7	c.211G>A (p.Gly71Arg)	+/+	Exon 1	c.1456T>G (p.Tyr486Asp)	+/+ (Exon 5)
		1/7		(TA)7/7	c.625C>T (p.Arg209Trp)	+/+	Exon 1	NR	NR
		1/7		(TA)7/7	c.686C>A (p.Pro229Gln)	+/-	Exon 1	NR	NR
Taiwan	Huang et al[33], 2006	1/3	0:1	NR	c.479T>A (p.Val160Glu)	+/+	Exon 1	NR	NR
		1/3	0:1	NR	c.610A>G (p.Met204Val)	+/+	Exon 1	NR	NR
		1/3	0:1	NR	c.1456T>G (p.Tyr486Asp)	+/+	Exon 5	c.211G>A (p.Gly71Arg)	+/-
Iran	Mohammadi Asl et al[20], 2013	8/12	4:4	NR	c.238_240delAG	+/+	Exon 1	NR	NR
					c.479T>A (p.Val160Glu)	+/+	Exon 1	NR	NR
GS
Africa	Horsfall et al[19], 2011	2616	NR	(TA)7/7, (TA)7/8	NR	NR	Exon 1	NR	NR
Sri Lanka	Premawardhena et al[34], 2003	229	115:114	(TA)7/7	NR	NR	NR	NR	NR
China	Wu et al[8], 2024	232/310	166:66	(TA)7/7, (TA)7/6, (TA)6/6	c.211G>A (p.Gly71Arg)	+/+, +/-	Exon 1	NR	NR
China	Gu et al[9], 2022	117	NR	(TA)7/7	c.211G>A (p.Gly71Arg)	+/+, +/-	Exon 1	NR	NR

M:F: Male:female ratio; CNS: Crigler-Najjar syndrome; NR: No record; +/+: Homozygous; +/-: Heterozygous; GS: Gilbert syndrome.

This study demonstrated an overall male predominance of 77.1%, which was consistent across the three types of UGT1A1-related disorders. However, the number of cases was limited, and the observed sex difference did not reach statistical significance; therefore, no causal inferences can be drawn. Similarly, Sun et al[30] reported a male predominance in 69.5% (n = 66/95) of their patients, including 43 with GS and 23 with CNS-II. A worldwide cohort study by Aronson et al[27], which included 221 patients with CNS from 13 countries, also reported a slight male predominance, with 54.3% (n = 120/221) males and 45.7% (n = 101/221) females. The single patient with CNS-I in our study was male. This finding is consistent with reports from Saudi Arabia[1] and Germany[14], where 66.6% (n = 8/12) and 75.0% (n = 9/12) of patients with CNS-I were male, respectively. In contrast, a study from the Netherlands reported a female predominance in CNS-I (n = 5/7, 71.4%)[15], whereas a study from Croatia reported an equal sex distribution among patients with CNS-I[28]. In the present study, 85.7% (n = 6/7) of patients with CNS-II were male, which is comparable to data from China, where 72.7% (n = 8/11) of patients with CNS-II were male[4]. Conversely, the Dutch study reported a female predominance in CNS-II (n = 6/9, 66.0%)[15]. However, both types of CNS are generally considered autosomal recessive disorders in which sex is not a determining factor[1,3,4-7,10,12]. CNS-II was historically considered to possibly follow an autosomal dominant inheritance pattern with incomplete penetrance or pseudodominance[1]. However, similar to CNS-I and GS, recent studies have described CNS-II as an autosomal recessive condition[4,5-7,10,12,13,17].

In this study, the single patient with CNS-I was genetically confirmed at the age of 6 months and 24 days. In contrast, the median age at diagnosis was 11.3 years (IQR: 0.3-16.8) for CNS-II and 13.4 years (IQR: 11.9-16.9) for GS. This diagnostic delay in CNS-II and GS may be attributed to their milder phenotypes compared with CNS-I[7] or to the coexistence of hematological disorders that can also increase indirect bilirubin levels, particularly G6PD deficiency (48.6%), SCD (40.0%), and alpha-thalassemia (14.3%). Inherited hematological disorders are common in Bahrain, with reported prevalences of 24.3% for alpha-thalassemia[35], 23.2% for G6PD deficiency, and 1.2% for SCD[29]. These conditions may receive greater clinical attention, potentially leading to reduced investigation for UGT1A1-related disorders and consequent underrecognition.

All patients in this study presented with persistent jaundice. The patient with CNS-I had neither hepatomegaly nor splenomegaly, whereas patients with CNS-II and GS exhibited hepatomegaly (14.3% and 22.2%, respectively) and splenomegaly (28.6% and 40.7%, respectively). Similarly, Kovačić Perica et al[28] reported no hepatomegaly or splenomegaly in three patients with CNS-I. In contrast, Li et al[4] reported that none of their patients with CNS-II had a palpable liver or spleen. GS typically does not present with organomegaly unless it is associated with coexisting hematological or hepatic disorders[36]. In the present study, all patients with CNS-II and GS who had hepatomegaly or splenomegaly also had SCD or thalassemia.

Beyond the physical manifestations of UGT1A1-related disorders, jaundice-related cosmetic concerns were reported in 45.7% of patients, who experienced emotional and social challenges that prompted them to seek medical advice. A similar observation was reported by Dhawan et al[11] in patients with CNS-I. Public misconceptions that jaundice is invariably caused by infectious diseases may lead to discrimination against patients with UGT1A1-related disorders. However, the psychosocial impact of these disorders warrants further investigation.

In this study, all patients had low hemoglobin levels and elevated reticulocyte percentages, suggesting hemolytic anemia. However, Erlinger et al[6] and Sun et al[30] reported that UGT1A1-related disorders are not typically associated with overt hemolysis. Similarly, Kovačić Perica et al[28] described eight patients with CNS-I, and Maruo et al[31] reported a case of CNS-II; in both reports, no evidence of hemolysis was observed. This discrepancy may be attributed to the high prevalence of hemolytic disorders in our cohort.

To date (April 7, 2026), the ClinVar database has identified 440 UGT1A1 gene variants[37]. Of these, 210 are missense mutations, 13 nonsense mutations, 34 frameshift mutations, and 14 variants located in the untranslated regions. Regarding variant types, 343 single-nucleotide substitutions, 46 deletions, 36 duplications, 22 insertions, and 4 indels have been reported[37]. Notably, the spectrum of UGT1A1 variants varies across different ethnic populations and countries (Table 6).

In this cohort, genetic testing of the patient with CNS-I revealed homozygosity for the (TA)7/7 promoter polymorphism along with the UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) mutation located in exon 3. These findings are consistent with those reported in two patients with CNS-I from unrelated Bedouin families in Kuwait[2]. In the Tunisian cohort, the (TA)7/7 genotype was also predominant, with 70.6% (n = 12/17) being homozygous; however, 29.4% (n = 5/17) were homozygous for the (TA)8/8 polymorphism[7]. In addition, all patients harbored the UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) mutation in exon 3 in a homozygous state[7]. The UGT1A1 (NM_000463.3):c.1070A>G (p.Gln357Arg) variant is rare and appears to be confined to Arab populations[2]. In contrast, the eight patients with CNS-I from Croatia had different genetic backgrounds[28]. Although homozygosity for (TA)7/7 was the predominant polymorphism (n = 3/7 tested patients, 37.5%), followed by (TA)6/6 (n = 2/7, 28.6%), the additional pathogenic variants differed[28].

In the present study, all seven patients with CNS-II carried the (TA)7/7 promoter polymorphism; 85.7% were homozygous, and 14.3% were compound heterozygous, in combination with the UGT1A1 (NM_000463.3):c.161G>A (p.Gly54Val) variant in exon 1 and other variants. Maruo et al[23] from Iran reported a patient with compound heterozygous variants, UGT1A1 (NM_000463.3):c.674T>G (p.Val225Gly) and c.381_382insGG (p.Cys127Trpfs*23), along with the A(TA)7TAA promoter variation. The UGT1A1 (NM_000463.3):c.161G>A (p.Gly54Val) variant has not been previously described in the literature. It affects a highly conserved residue, and bioinformatic analysis predicts a deleterious effect. Based on the currently available evidence, this variant should be classified as a variant of uncertain significance (VUS). Similarly, the UGT1A1(NM_000463.3):c.907G>A (p.Val303Met) variant has been classified in ClinVar as a VUS[38]. To date, no associations between the UGT1A1 (NM_000463.3):c.161G>A (p.Gly54Val) or the UGT1A1 (NM_000463.3):c.907G>A (p.Val303Met) variants and CNS have been reported in the literature. Further studies are required to investigate potential associations with CNS and to evaluate their functional impact on UGT1A1 enzyme activity before considering them novel pathogenic variants.

For GS, the most prevalent variant in this study was homozygosity for the (TA)7/7 promoter polymorphism in the UGT1A1 gene, with no additional variants detected (n = 24/27, 88.9%). Homozygosity for (TA)7/7 in the promoter region without additional pathogenic variants is the most frequently reported genetic cause of GS[4,8,19,34]. Gu et al[9] evaluated 117 Chinese patients with GS and reported that the most common pathogenic variants were the promoter A(TA)7TAA insertion and the p.Gly71Arg missense variant. Only two patients with GS (7.4%) in our cohort were compound heterozygous for (TA)7/(TA)8, whereas one patient harbored an additional variant. The (TA)8 allele is extremely rare in Caucasian populations[39]. Iolascon et al[39] reported a case of a girl with GS from Taranto (Apulia), Italy, who was compound heterozygous for (TA)7/(TA)8. Coelho et al[40] also described a 5-year-old Portuguese Caucasian girl with GS who was compound heterozygous for two promoter insertions, (TA)7 and (TA)8.

Management of UGT1A1-related disorders varies depending on the patient’s clinical condition. Patients with CNS-I may require exchange transfusions, long-term phototherapy, or LTx, whereas those with CNS-II are typically managed with lifelong phenobarbital therapy[1-4,10,11,14,15,20,27]. Patients with GS generally do not require medical intervention because of its milder phenotype, with serum bilirubin levels usually ranging from normal to 85 μmol/L[33]. In the present study, the single patient with CNS-I initially received hospital-based phototherapy combined with phenobarbital before diagnostic confirmation. He subsequently continued home phototherapy until the age of 3.7 years, when he underwent auxiliary living-related LTx, which resulted in rapid normalization of serum bilirubin levels. Sambati et al[17] noted that management of CNS-I requires intensive interventions, including phototherapy and plasmapheresis, whereas phenobarbital, which induces UGT1A1 enzyme expression, is ineffective because of the complete absence of functional UGT1A1 activity in CNS-I[5]. Whole-body phototherapy should be initiated as early as possible after birth[41], ideally following exchange transfusion if indicated[1,11]. Adherence to phototherapy imposes a substantial burden on both patients and caregivers[11]. CNS-I has a significant emotional impact on children, as they face limitations in social interactions and may feel stigmatized because of their appearance[11]. Moreover, caregivers and parents of children with CNS-I experience financial, psychological, and emotional hardship due to the demands of continuous phototherapy and the requirement to keep toddlers under blue light exposure for 10-12 hours daily[3,11,16,26,41]. This was also observed in our patient with CNS-I before LTx. Moreover, phototherapy becomes less effective with increasing age because of increased skin pigmentation and thickness, a less favorable surface area-to-body weight ratio, limited hepatobiliary clearance of lumirubin, and decreased adherence[12,16]. To enhance the effectiveness of nocturnal phototherapy, ursodeoxycholic acid (15-30 mg/kg/day) and a lipid-rich bedtime snack have been recommended, as they may stimulate bile flow and promote hepatic clearance of lumirubin[17]. However, LTx remains the only definitive treatment, particularly when phototherapy and plasmapheresis are no longer effective[17]. Two main LTx approaches are commonly performed: Orthotopic LTx (OLT) and auxiliary partial OLT[11]. In the largest cohort reported by Aronson et al[27], 11.8% (n = 26/221) of patients with CNS underwent LTx at a median age of 9 years (range: 0-32), resulting in a significant reduction in total bilirubin levels. In Croatia, LTx was performed in 50% of patients with CNS-I (n = 4/8) at the ages of 5.5, 7, 9.5, and 10 years[28]. Although the optimal timing of LTx in patients with CNS-I remains controversial, transplantation should be performed before the onset of neurological complications[12]. The recommended age for LTx in children with CNS-I is between 3 and 5 years, ideally before school entry[12]. Beyond this age, the effectiveness of phototherapy decreases substantially, increasing the risk of neurological injury[12]. Ozçay et al[3] reported four patients with CNS-I who underwent OLT at the ages of 2 months, 8.5 months, 15 months, and 13 years; three demonstrated rapid normalization of bilirubin levels, survived, and achieved normal neurodevelopmental milestones. A similar outcome was observed in our patient with CNS-I, who experienced rapid resolution of jaundice after LTx. Although LTx remains the gold standard treatment for CNS-I, novel therapeutic strategies, such as hepatocyte transplantation and gene therapy, warrant further investigation.

In the present study, 85.7% (n = 6/7) of patients with CNS-II received phenobarbital therapy. Phenobarbital induces UGT1A1 gene transcription, thereby increasing UDP-GT activity and reducing plasma unconjugated bilirubin concentrations by 30% or more[12,41]. Phenobarbital is recommended as long-term therapy for patients whose serum bilirubin levels exceed 15 mg/dL or for those with jaundice that affects their quality of life[42]. The recommended dose is 2 mg/kg administered two to three times daily for children or 60-180 mg/day for adults, divided into two doses[42]. Our study demonstrated a modest reduction in mean indirect bilirubin levels following phenobarbital therapy. Nonetheless, adherence to lifelong medication remains challenging, particularly among adolescents[11]. Adherence to phenobarbital was poor in 83.3% of patients with CNS-II. The reasons for poor adherence included patient refusal, parental reluctance to administer the medication, and discontinuation of therapy following consultation with adult gastroenterologists. Moreover, in our hospital, caregivers are required to attend outpatient clinics every two months to refill phenobarbital prescriptions, a policy that may negatively affect adherence and, consequently, treatment outcomes. In addition, the use of lower-than-recommended doses of phenobarbital may reduce therapeutic effectiveness. Social stigma associated with the use of anticonvulsant medication was also observed in our cohort, contributing to parental refusal to initiate therapy. However, the clinical manifestations of CNS-II are heterogeneous; therefore, not all patients require continuous phenobarbital therapy, and treatment may be periodically interrupted in selected cases[5].

Although patients with GS generally do not require treatment[9], phenobarbital is occasionally used[5], as the potential adverse effects of the medication may outweigh the benefits of treating such a benign degree of jaundice. In our study, 48.1% (n = 13/27) of patients with GS received phenobarbital therapy to reduce the cosmetic impact of jaundice before diagnosis while awaiting genetic test results. Historically, phenobarbital administration was used as a diagnostic test to confirm GS before the availability of genetic testing, to distinguish GS from CNS-I and CNS-II. Other diagnostic approaches included fasting bilirubin measurement, rifampin administration, and assessment of UGT enzyme activity via thin-layer chromatography, after excluding hepatobiliary and hemolytic disorders[9,26].

In the present study, 22.9% (n = 8/35) of patients with UGT1A1-related disorders underwent cholecystectomy for the treatment of cholelithiasis. However, it remains unclear whether cholelithiasis was attributable to hemolytic anemia secondary to associated hematological disorders or to UGT1A1-related disorders. Strauss et al[26] reported that 43% (n = 12/28) of patients with CNS-I were diagnosed with cholelithiasis; all underwent cholecystectomy, which revealed multiple small pigment stones.

Several limitations should be acknowledged. The retrospective, single-center design resulted in some missing data and limits the generalizability of the findings. However, this center serves as the national and sole referral center for genetic services in Bahrain, receiving referrals from other tertiary hospitals; therefore, the cohort likely represents genetically confirmed UGT1A1-related disorder cases evaluated nationwide. Nonetheless, prevalence estimates should be interpreted as minimum estimates and may not fully capture undiagnosed or untested individuals. In addition, the clustering of diagnoses in recent years likely reflects detection bias related to improved access to genetic testing rather than a true increase in disease incidence. Although the study spans two decades, the marked increase in case identification during the most recent five-year period suggests substantial underdiagnosis in earlier years, which may have affected the accuracy of prevalence estimates. Another limitation is the relatively small sample size, reflecting the rarity of UGT1A1-related disorders and referral-based case ascertainment. Moreover, because molecular testing was limited to the UGT1A1 gene, undetected pathogenic variants in other genes involved in bilirubin metabolism may have contributed to hyperbilirubinemia in a subset of patients. The classification of certain patients, particularly those with CNS-II or GS, was challenging because of overlapping genetic findings, including VUS, and the frequent coexistence of hemolytic disorders. Furthermore, phenobarbital was initiated in a subset of patients as a therapeutic trial prior to genetic testing; subsequent confirmation of GS may have influenced serum bilirubin levels at follow-up. Moreover, neither the UDP-GT enzyme assay nor the pyruvate kinase enzyme assay was performed. In addition, detailed documentation regarding phototherapy management and its duration in patients with CNS-II and GS was lacking. Similarly, the exact phenobarbital dose, dose variability, and duration of therapy, as well as the detailed psychological impact of jaundice and patient-reported treatment experiences, were not recorded. Finally, fluctuations in unconjugated bilirubin levels, which may occur during acute infections or rebound hyperbilirubinemia, were not analyzed.

Despite these limitations, the findings of this study are clinically relevant. This is the first study from Bahrain to address UGT1A1-related disorders, encompassing key aspects ranging from disease prevalence and genetic characterization to clinical outcomes, including all three phenotypes. The results may assist healthcare providers in the diagnosis and management of patients with UGT1A1-related disorders. These findings may also provide a foundation for future research exploring various aspects of this rare condition. In addition to the physical impact of the disease, physicians should address the mental and psychological well-being and quality of life of patients with UGT1A1-related disorders and their caregivers to ensure comprehensive care.

CONCLUSION

This study found that UGT1A1-related disorders are an important cause of persistent indirect hyperbilirubinemia in Bahrain. Among these disorders, GS was the most common phenotype, followed by CNS-II and CNS-I. Despite its higher prevalence, GS likely represents only the visible portion of the burden, as it remains largely underdiagnosed because of its typically mild and often unnoticed clinical manifestations. GS is frequently misdiagnosed as CNS-II, a less common condition, which may lead to unnecessary treatment with phenobarbital. Although extremely rare, CNS-I has severe clinical consequences because of markedly elevated bilirubin levels and the associated risk of kernicterus. Management often involves prolonged phototherapy, which can substantially impair quality of life, and LTx, which carries considerable risks. Diagnostic evaluation for UGT1A1-related disorders, including enzyme assays and comprehensive genetic sequencing, should be prioritized in patients with inherited hematological disorders, particularly those with G6PD deficiency and SCD, especially in populations with high rates of consanguinity, to minimize delays in diagnosis. Further research is needed to develop novel therapeutic approaches for CNS-I, such as hepatocyte transplantation and gene therapy. In addition, future studies should evaluate phenobarbital treatment strategies (continuous vs intermittent administration), treatment adherence in patients with CNS-II, and the psychosocial impact and long-term outcomes in individuals with GS.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge all healthcare professionals who provided care for patients with genetic and hematological conditions in the Department of Pediatrics, Salmaniya Medical Complex, Government Hospitals, Kingdom of Bahrain.