Neonatal Jaundice — Genetic Forms: Crigler–Najjar Syndrome and Gilbert Syndrome

Inherited UGT1A1-related unconjugated hyperbilirubinaemias, with the differential diagnosis of other genetic causes of jaundice in newborns and infants.

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

Newborn jaundice is usually a transient physiological adaptation: unconjugated, or indirect, hyperbilirubinaemia appears on the second or third day and resolves spontaneously. Investigation is necessary when jaundice persists beyond two to four weeks, appears in the first 24 hours, approaches phototherapy thresholds or involves an elevated conjugated, or direct, fraction, because neonatal cholestasis is always pathological [1, 13, 14]. The geneticist is usually consulted after common paediatric causes have been excluded to determine whether hyperbilirubinaemia has a genetic cause or contribution [1]. This article addresses inherited unconjugated hyperbilirubinaemias caused by UGT1A1 variants — Crigler–Najjar syndrome type 1 (OMIM 218800), type 2 (OMIM 606785) and Gilbert syndrome (OMIM 143500) — and incorporates other genetic causes of neonatal and childhood jaundice, both haemolytic and cholestatic, including glucose-6-phosphate dehydrogenase (G6PD) deficiency, Dubin–Johnson and Rotor syndromes and familial intrahepatic cholestases [1, 5, 7].

Gilbert syndrome was the first condition to be described. In 1900, Gilbert, Castaigne and Lereboullet reported asymptomatic serum bilirubin elevation without evident haemolytic disease under cholémie simple familiale; the condition was subsequently studied by van den Bergh, who developed the bilirubin diazo assay still used in clinical chemistry, and by the Danish physician Jens Einar Meulengracht, whose name joined Gilbert’s in German-language literature (Gilbert–Meulengracht syndrome) [9]. Names such as icterus intermittens juvenilis, hereditary non-haemolytic bilirubinaemia and constitutional hepatic dysfunction were abandoned, and Gilbert syndrome, or benign hyperbilirubinaemia, is now the only term used [9].

In 1952 John Crigler and Victor Najjar described “congenital familial non-haemolytic jaundice with kernicterus” in Pediatrics, in seven infants from three families who developed intractable non-haemolytic jaundice during their first week of life [16, 28]. Six died of bilirubin encephalopathy by 15 months, defining the natural history of the severe form before phototherapy became available [12, 16]. In 1969 Arias and colleagues assembled clinical, biochemical, pharmacological and genetic evidence for heterogeneity in chronic non-haemolytic hyperbilirubinaemias with glucuronyltransferase deficiency, separating an intermediate-severity form responding to enzyme induction from Crigler and Najjar’s form; this became Crigler–Najjar syndrome type 2 or Arias syndrome [7, 29].

The shared molecular basis emerged in subsequent decades. UDP-glucuronosyltransferase 1A1 (UGT1A1) is the only isoform significantly contributing to bilirubin conjugation, and genetic lesions inactivate it completely in type 1 and partially in type 2 [30]. In 1995 Bosma and colleagues demonstrated that patients with Gilbert syndrome had normal coding regions but were homozygous for two additional nucleotides (TA) in the promoter TATA element, A(TA)7TAA instead of A(TA)6TAA, which reduced reporter gene expression and was necessary but insufficient for full expression of the syndrome [8]. These three conditions have since been viewed as a UGT1A1 deficiency spectrum, with residual enzyme activity determining bilirubin concentration, neurological risk and treatment [2, 31].

Phototherapy changed the prognosis of the severe form: before its discovery, every patient without residual activity died shortly after birth, whereas optimised equipment enabled prevention of brain injury and survival until liver transplantation [12]. A registry published in 1996 identified orthotopic liver transplantation and daily phototherapy as the two principal strategies. The 2022 worldwide registry of 221 patients showed that these remain the basis of treatment, with regional survival differences depending mainly on access to high-quality phototherapy [12]. Liver-directed gene therapy is undergoing clinical evaluation and is discussed in section 7.

Gilbert syndrome has a second, pharmacogenetic dimension. Reduced UGT1A1-mediated glucuronidation alters the clearance of medicines such as irinotecan and explains atazanavir-induced jaundice; Clinical Pharmacogenetics Implementation Consortium (CPIC) and Dutch Pharmacogenetics Working Group (DPWG) guidelines exist for these gene–drug pairs [9, 21, 22]. The article consistently distinguishes Mendelian disease, common predisposition and the pharmacogenetic effect, because mixing these planes leads to counselling errors.

2. Molecular Genetics and Pathophysiology

2.1. Bilirubin metabolism and inherited block sites

Bilirubin arises mainly from haem breakdown in senescent erythrocyte haemoglobin. Haem oxygenase converts haem to biliverdin, and biliverdin reductase A reduces this to lipophilic unconjugated bilirubin, which circulates bound to albumin and is transported to the liver [1, 2]. In the smooth endoplasmic reticulum of hepatocytes, UGT1A1 conjugates bilirubin with glucuronic acid into water-soluble mono- and diglucuronides; these are excreted into bile canaliculi by multidrug resistance-associated protein 2 (MRP2, encoded by ABCC2). When canalicular secretion saturates, conjugates may return to sinusoids through MRP3 and be recaptured by downstream hepatocytes through organic anion transporting polypeptides OATP1B1 and OATP1B3, encoded by SLCO1B1 and SLCO1B3 [2]. Unconjugated bilirubin is not renally excreted and is toxic to the brain, potentially causing kernicterus and deafness [1].

This pathway locates the entities covered in this article. Conjugation deficiency caused by UGT1A1 variants produces predominantly unconjugated hyperbilirubinaemia: Gilbert syndrome and Crigler–Najjar syndrome types 2 and 1. Canalicular excretion deficiency caused by biallelic ABCC2 variants produces Dubin–Johnson syndrome, while simultaneous absence of OATP1B1 and OATP1B3 through biallelic SLCO1B1 and SLCO1B3 variants produces Rotor syndrome; in these two conditions, hyperbilirubinaemia is predominantly conjugated [2, 7]. In newborns, haemolytic overproduction combines with immature conjugation: erythrocyte lifespan is shorter, UGT1A1 activity is low and enterohepatic bilirubin circulation is increased [2].

2.2. The UGT1A1 gene, protein and variant nomenclature

UGT1A1 is located at 2q37.1 (gene OMIM 191740; HGNC:12530; 2q37 deletions encompassing 2q37.1 remove one copy of the gene) and comprises an enhancer, a promoter and five exons [2, 31]. MANE Select transcript NM_000463.3 encodes protein NP_000454.1, confirmed in VariantValidator [4]. UGT1A1 is the only human enzyme that makes a substantial contribution to bilirubin glucuronidation and also conjugates oestrogens, arachidonic acid metabolites and medicines such as SN-38, irinotecan’s active metabolite [22, 30]. In January 2025, the Human Gene Mutation Database recorded at least 189 pathogenic UGT1A1 variants [2].

Literature uses star allele nomenclature inherited from pharmacogenetics, older protein designations and Human Genome Variation Society (HGVS) nomenclature simultaneously. Promoter TA repeats are the most demanding example: *28 is historically described as A(TA)7TAA or c.-41_-40dupTA on NM_000463.2, but ClinVar expresses it as genomic repeat NC_000002.12:g.233760235TA[8] because it counts the final TA unit in TAA. By the same convention, *36 (A(TA)5TAA) corresponds to g.233760235TA[6] and *37 (A(TA)8TAA) to g.233760235TA[9] [32]. VariantValidator rejects c.-41_-40dup on NM_000463.3 as outside the transcript reference sequence and returns genomic NC_000002.12:g.233760247_233760248dup [4]. Reports should therefore state the number of repeats, historical designation and genomic description without manual numbering conversion.

The last row and the p.Glu241GlyfsTer16 row illustrate two practical problems. In the first, the same enhancer variant appears with two numbering systems in sources from 2025 and the transcript description is rejected by VariantValidator; laboratory designation and genomic position should be cited. In the second, the Croatian series reported c.717_718delAG and c.722_723delAG as separate mutations, but VariantValidator right-aligns the former and returns the latter with the identical protein consequence p.(Glu241GlyfsTer16) [4, 11]. This observation arises from the normalisation performed for this article and was not discussed by the authors; it matters for interpreting a possible founder effect.

2.3. Enzyme deficiency degree and genotype–phenotype correlation

Clinical severity follows residual UGT1A1 activity. In Gilbert syndrome, hepatic glucuronidation activity is reduced to approximately 30 % of normal [2, 8]; the Croatian series gives 20 to 30 % in Gilbert, 4 to 10 % in type 2 and no activity in type 1, while the Chinese consensus defines type 2 as below 10 % [2, 11]. For *28, heterozygosity reduces activity by approximately 10 to 35 % and homozygosity by approximately 70 % [2]. Strauss and colleagues note that 30 % or less of normal hepatic UGT1A1 activity suffices for bilirubin homeostasis, which explains the effectiveness of partial grafts and the objectives of cellular and gene therapies [16]. These values derive from functional studies and clinical series; they explain the spectrum rather than predict an individual patient’s bilirubin concentration.

Type 1 results from variants that abolish function — deletions, insertions, nonsense, frameshift, splice variants or critical residue substitutions — in homozygous or compound heterozygous form [2, 11]. Type 2 generally results from compound heterozygosity for partial loss-of-function variants or a Crigler–Najjar-type variant combined in trans with a Gilbert-type promoter insertion [1, 30]. Intermediate hyperbilirubinaemia levels are therefore common in Crigler–Najjar relatives [1]. Bosma and colleagues observed in a type 2 family that only six heterozygotes who carried the long promoter on the structurally normal allele had mild hyperbilirubinaemia compatible with Gilbert syndrome [8].

In Gilbert syndrome, *28 homozygosity is necessary in most European cases but insufficient, and penetrance depends on bilirubin production, sex and environmental factors [1, 30]. Expression is much less frequent in adult women, possibly from lower daily bilirubin production and sex steroid modulation of UGT1A1, although *28 genotype frequency is similar in both sexes [1, 9]. In Asian populations, where *28 is less common, heterozygous coding variants such as *6 predominate [7, 9]. Reduced hepatic bilirubin uptake has also been described in Gilbert syndrome, but the responsible transporter is unidentified and no genetic studies of this component exist [9]. The boundaries between Gilbert syndrome and type 2 are indistinct: bilirubin fluctuates within individuals and phenotypes overlap, but unconjugated hyperbilirubinaemia with homozygous A(TA)7TAA and no pathogenic coding variants should be diagnosed as Gilbert syndrome [7].

2.4. Bilirubin toxicity and neurological injury

Free unconjugated bilirubin, unbound to albumin, crosses the blood–brain barrier and is deposited in the basal ganglia and brainstem nuclei [2]. Acute bilirubin encephalopathy begins with hypotonia, lethargy, high-pitched cry and poor sucking and may progress to hypertonia, opisthotonus, irritability, fever, seizures and apnoea; chronic kernicterus leaves extrapyramidal motor dysfunction, sensorineural deafness, ocular motility abnormalities and enamel hypoplasia [2]. Risk increases above a bilirubin/albumin molar ratio of 0.8, which explains the attention paid to albumin, acidosis and sepsis, which increase the free fraction [2, 14]. In the worldwide registry, unconjugated bilirubin concentrations up to approximately 300 µmol/L were associated with encephalopathy risk [12].

2.5. Photoisomerisation and therapeutic mechanisms

Phototherapy uses blue light near 450–478 nm and converts unconjugated bilirubin to water-soluble photoisomers excreted in bile and urine without hepatic conjugation [2, 12]. Effectiveness decreases with age through increased skin thickness and pigmentation and reduced surface area-to-weight ratio [2, 11]. Phenobarbital increases UGT1A1 transcription and reduces bilirubin only when residual activity is present; it is ineffective when the enzyme is absent [12]. Liver transplantation replaces the enzyme in the required organ, and hepatic gene therapy seeks the same outcome using a viral vector [2, 12].

2.6. Related diseases: conjugate transporters

In Dubin–Johnson syndrome, loss of MRP2 prevents canalicular export of conjugates, causing reflux into plasma; this is a transport defect with intact upstream conjugation [7, 34]. In Rotor syndrome, the simultaneous absence of OATP1B1 and OATP1B3 prevents recapture of bilirubin glucuronides exported into the blood and reduces hepatic uptake of anionic dyes and cholescintigraphic radiopharmaceuticals; hyperbilirubinaemia requires biallelic inactivating variants in both genes, because one functional allele of either gene prevents it [7]. The genes are close together at 12p12, so digenic inheritance behaves clinically as monogenic autosomal recessive inheritance [7].

2.7. Inheritance

Crigler–Najjar syndrome types 1 and 2 are autosomal recessive [5, 30]. Gilbert syndrome usually represents homozygosity for a common promoter allele or combinations of reduced-function alleles. Some classifications treat it as a polymorphism, the Chinese consensus treats it as an autosomal recessive disorder, and Orphanet considers it non-rare in Europe, outside its rare disease nomenclature [2, 6]. Dubin–Johnson syndrome is autosomal recessive, while Rotor syndrome is digenic with a recessive-like segregation pattern [7, 34].

3. Clinical Spectrum and Manifestations by System

3.1. Overview

The three UGT1A1-associated conditions share unconjugated hyperbilirubinaemia without haemolysis or primary hepatocellular injury, but differ in bilirubin concentration, age at recognition, phenobarbital response and neurological risk [1, 2]. Bilirubin ranges differ slightly between sources because they were defined in different populations and periods; the following table presents them alongside each other without merging them.

3.2. Crigler–Najjar type 1

Jaundice appears in the first days of life, worsens over subsequent weeks and is not accompanied by pale stools, dark urine, hepatomegaly or splenomegaly [1, 11]. In newborns with type 1 not yet receiving phototherapy, total bilirubin rose by 4.3 ± 1.1 mg/dL daily (range 2.2–6.0), reaching potentially neurotoxic concentrations between 5 and 14 days in a US cohort of 28 p.(Tyr74Ter) homozygotes [16]. Dhawan and colleagues’ systematic review describes a neonatal rise of 3 to 6 mg/dL daily [35]. Endogenous bilirubin production, 3.7 ± 0.9 mg/kg daily, resembles that of healthy adults, so hyperbilirubinaemia reflects deficient elimination rather than overproduction [16].

In the same cohort, four of 28 children (14 %) developed kernicterus between 14 and 45 days of life, with total bilirubin concentrations of 30 to 48 mg/dL (513–821 µmol/L) and a bilirubin/albumin molar ratio of 1.10 to 1.34; the sequelae included dystonia–chorea and dysphagia in all four, mutism in all four and hearing impairment in two [16]. The children who developed kernicterus began phototherapy at a median age of 37 days (14–91), compared with 4 days (1–27) in the remaining children [16]. Neurological injury may occur beyond the neonatal period: in the Croatian series, a child whose phototherapy was discontinued at 9 months developed regression of development and language, postural instability and choreoathetosis at 5 years [11]. The 2022 worldwide registry reports an overall incidence of neurological symptoms of 10.7 %, but 42.7 % in the subgroup of 75 patients without adequate access to phototherapy, with generalised hypertonia and dystonia, hypotonia, cerebellar dysfunction and dysphasia [12]. The signs of acute and chronic encephalopathy are described in section 2.4; Saudubray includes kernicterus with athetosis among the extraneurological and neurological signs that point towards Crigler–Najjar syndrome when assessing inherited metabolic diseases [2, 36]. Prolonged hyperbilirubinaemia above 250 µmol/L is one of the environmental factors associated with deafness in early childhood [37].

Other hepatic functions are preserved, and the liver was traditionally considered morphologically normal [1], but this view has been revised. In the worldwide registry, alanine aminotransferase (ALT) exceeded 70 U/L or twice the upper limit in 43 of 113 patients with data (38.1%), without correlation with bilirubin. Advanced fibrosis was documented in 5 of 31 assessed patients (16%) and in 5 of 15 (33%) phototherapy-dependent patients aged 7 or more [12]. In Strauss’s cohort, 9 of 15 examined livers (60%) had mild to severe fibrosis, correlated with lifetime mean bilirubin [16]. Gallstones were found in 14 of 91 ultrasound examinations (15.4%) in the worldwide registry and 12 of 28 patients (43%) in the Amish and Mennonite cohort, all of whom underwent cholecystectomy for pigment stones [12, 16]. Total bilirubin was higher in males (316.0 ± 125.3 µmol/L) than females (286.6 ± 116.6 µmol/L) in the worldwide registry [12].

The skin of patients receiving prolonged phototherapy may develop xerosis, lichenification, hyperpigmentation and, rarely, an erythematous eruption; one patient in the registry developed melanoma, without an established causal relationship with phototherapy [12]. In the North American cohort, five patients with more than 100,000 hours of exposure without eye protection had normal visual acuity and colour discrimination, and neither erythema nor skin cancer was observed [16].

3.3. Crigler–Najjar type 2

Type 2 presents with persistent jaundice of intermediate severity, often recognised in childhood, and generally has a favourable prognosis [2]. The risk of kernicterus is lower than in type 1, but is not zero [5]: cases of bilirubin encephalopathy triggered by trauma or surgical stress have been described in adults, particularly when the bilirubin/albumin molar ratio exceeds 0.8 [2]. In a 2026 report, an adolescent with type 2 developed bilirubin-induced neurological dysfunction and was treated with therapeutic plasma exchange [38]. UGT1A1 variants are associated with an increased risk of gallstones, and some patients with type 2 develop fibrosis or cirrhosis through mechanisms that remain unclear [2].

Junge and colleagues’ German series showed that a specific genotype may combine a mild bilirubin phenotype with severe liver disease. Fourteen patients homozygous for c.115C>G p.(His39Asp), almost all from Lebanese and Turkish families and all those analysed also homozygous for *28, required no regular phototherapy. Seven developed severe fibrosis or cirrhosis, six underwent transplantation and two died, one without transplantation and one afterwards [10]. The cause of liver disease was not identified, and the authors question whether Crigler–Najjar fibrosis depends solely on bilirubin or phototherapy [10].

3.4. Gilbert syndrome

Gilbert syndrome causes mild, fluctuating unconjugated hyperbilirubinaemia, frequently discovered during routine testing or after intercurrent illness in adolescents and young adults [1, 9]. Jaundice is triggered or aggravated by fasting, dehydration, excessive fatigue, psychological stress, the menstrual cycle, infections and puberty [2]. Hepatosplenomegaly and other signs of liver disease are absent [2]. Non-specific symptoms previously attributed to the syndrome, including fatigue, asthenia and dyspepsia, are no longer considered part of it and require investigation [9]. Erythrocyte lifespan may be shortened by approximately 30% without evident haemolysis [9].

The clinical associations described by Vítek and Tiribelli include risks and possible benefits. The risks are drug interactions, more intense neonatal jaundice, a more severe course of inherited haemolytic diseases and a higher prevalence of pigment gallstones. The associated benefits, derived mainly from observational studies, are a lower risk of atherosclerotic disease, diabetes, cancer, autoimmune and neurodegenerative diseases, and a lower body mass index at older ages [9]. In a United Kingdom primary care database, mortality was almost 50 % lower in people with Gilbert syndrome, but these data do not establish causality [9].

3.5. UGT1A1 polymorphisms and newborn jaundice

Outside Mendelian disease, reduced-function alleles modulate neonatal jaundice through interactions with other factors. Prolonged neonatal jaundice is found in newborns who are heterozygous or homozygous for the Gilbert variant and also have G6PD deficiency, Coombs-negative ABO disease, spherocytosis or prolonged breastfeeding [1]. In a study of term newborns screened for Mediterranean-type G6PD deficiency, total bilirubin ≥257 µmol/L occurred in 22.9 % of G6PD-deficient infants and 9.2 % of controls. Among the deficient infants, the frequency was 9.7 % with a normal promoter genotype, 31.6 % in heterozygotes and 50 % in homozygotes for the variant promoter, whereas the polymorphism had no significant effect in newborns without G6PD deficiency [39]. The association of *28 alone with neonatal hyperbilirubinaemia is inconsistent: a meta-analysis of 13 studies found an association with TA7/TA7 homozygosity in Asia and Europe, and with p.Gly71Arg in Asia and Africa, whereas another analysis, including four studies, found no significant association with *28 [40, 41]. Breast milk jaundice may prolong unconjugated hyperbilirubinaemia for up to three months and is almost always non-pathological [14]. The American Academy of Pediatrics (AAP) guideline also lists hypertrophic pyloric stenosis associated with Gilbert syndrome among the causes of prolonged indirect hyperbilirubinaemia [14].

3.6. Pregnancy in Crigler–Najjar patients

Pregnancy in a woman with Crigler–Najjar syndrome places the fetus at risk because maternal unconjugated bilirubin crosses the placenta [2, 15]. Management is described mainly in case reports. A 2009 systematic review of type 2 reported that maternal treatment with phenobarbital reduced unconjugated bilirubin and prevented fetal and neonatal sequelae [42]. A successful pregnancy has also been reported in a patient with type 1 treated with phototherapy and fortnightly albumin infusions [43]. The review by Sambati and colleagues describes two children of mothers with type 1 who required exchange transfusion, one of whom had sensorineural deafness [15]. The bilirubin and bilirubin/albumin ratio targets proposed for pregnancy are presented in section 7.

4. Clinical Recognition and Warning Signs

Recognition begins by distinguishing physiological jaundice from jaundice that requires investigation. The two major neonatal guidelines define different time limits, which should be considered together. The National Institute for Health and Care Excellence (NICE) considers jaundice prolonged when it lasts more than 14 days in newborns of 37 or more weeks of gestation and more than 21 days in those of lower gestational age [13]. The AAP recommends measuring total bilirubin and the direct or conjugated fraction in breastfed infants who remain jaundiced at 3–4 weeks and formula-fed infants who remain jaundiced at 2 weeks [14]. Joint guidance from the North American and European societies for paediatric gastroenterology, hepatology and nutrition (NASPGHAN and ESPGHAN) recommends assessing cholestasis in any infant who is jaundiced after 2 weeks of age [44]. The AAP adds that an infant aged 7 or more days whose total bilirubin remains less than 2 mg/dL below the phototherapy threshold may have prolonged indirect hyperbilirubinaemia. Most have breast milk jaundice, but the list of causes includes haemolytic disease, hypothyroidism, extravascular blood, pyloric stenosis with Gilbert syndrome and Crigler–Najjar syndrome [14].

The history taken by the geneticist should include consanguinity, ethnic origin, prolonged neonatal jaundice in siblings or parents, reactions to medicines and foods, other structural anomalies such as congenital heart disease, an affected half-sibling on the maternal side, intrahepatic cholestasis of pregnancy in the mother, the onset and fluctuations of jaundice, pruritus, previous treatments and poor weight gain with steatorrhoea [1]. Physical examination should assess neurological abnormalities, a large fontanelle, facial dysmorphism, cataracts, hepatosplenomegaly, a heart murmur, bruising and bleeding, contractures, and the colour of urine and stools [1]. The absence of dysmorphism, organomegaly and changes in stool colour, together with markedly elevated indirect bilirubin, points towards Crigler–Najjar syndrome [1, 11].

Recognition has different consequences in each group. In type 1, starting phototherapy after day 13 was associated with a 3.5-fold greater risk of kernicterus [16]. In Gilbert syndrome, the principal danger is diagnostic and pharmacological: unnecessary liver investigations and toxicity from glucuronidated medicines [9]. Coexisting Gilbert syndrome should be considered in controlled liver disease, such as chronic hepatitis B with undetectable viral DNA, when isolated indirect hyperbilirubinaemia persists after haemolysis is excluded [2].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Measuring and fractionating bilirubin

Serum total bilirubin is the definitive test guiding decisions about phototherapy and escalation of treatment in newborns (AAP key action statement 3). Transcutaneous bilirubin may be used for the initial assessment, but serum total bilirubin should be measured when the transcutaneous value is within 3 mg/dL of the phototherapy threshold or is equal to or greater than 15 mg/dL (key action statement 6), and the direct or conjugated fraction should not be subtracted from the total when applying the thresholds [14]. “Direct” and “conjugated” bilirubin are not equivalent: the former includes a small fraction of unconjugated bilirubin that reacts without an accelerator, is more variable and tends to rise with the total [14]. NASPGHAN/ESPGHAN guidance considers direct bilirubin above 1.0 mg/dL (17 µmol/L) abnormal, and other authors propose investigation from a conjugated bilirubin concentration of 0.3 mg/dL (5 µmol/L); a proportion above 20 % of the total is no longer required to define cholestasis [14, 44, 45]. NICE requests specialist advice when conjugated bilirubin exceeds 25 µmol/L [13]. In patients with Crigler–Najjar syndrome, conjugated bilirubin may be falsely elevated because of interference from very high concentrations of unconjugated bilirubin, delta bilirubin covalently bound to albumin or photoisomers. In the worldwide registry, mean conjugated bilirubin was 14.4 µmol/L and did not correlate with transaminases [12]. This interference should neither divert investigation towards nonexistent cholestasis nor, conversely, lead to genuine cholestasis being overlooked.

5.2. Excluding haemolysis and other common causes

In infants with prolonged jaundice, NICE recommends looking for pale stools and dark urine, measuring conjugated bilirubin, obtaining a full blood count, determining the blood groups of the mother and newborn and performing a direct antiglobulin test (Coombs), considering urine culture if urinary tract infection is suspected, and confirming that metabolic screening, including screening for congenital hypothyroidism, has been performed [13]. The AAP recommends reviewing newborn screening results because galactosaemia, hypothyroidism and tyrosinaemia can cause persistent jaundice, and measuring G6PD activity in any newborn with jaundice of unknown cause whose bilirubin rises despite intensive phototherapy, rises suddenly or rises again after falling, or who requires escalation of care [14]. G6PD activity measured during or shortly after a haemolytic crisis or exchange transfusion may be falsely normal and should be measured again at least three months later [14].

Haemolysis is recognised by reticulocytosis (above 3–4 % or more than 120 × 10⁹/L), a fall in haemoglobin, abnormalities on the blood film, elevated lactate dehydrogenase, reduced haptoglobin and the Coombs test, which is positive in immune forms [2]. In UGT1A1 disorders, the full blood count is normal, reticulocytes are usually below 1.5 %, the blood film is normal and the Coombs test is negative. However, Gilbert or Crigler–Najjar syndrome may coexist with a haemolytic disorder, and it is precisely this combination that produces the most intense hyperbilirubinaemia [2, 39].

5.3. Gilbert syndrome diagnosis: two approaches

The review by Vítek and Tiribelli, published in the Journal of Hepatology in 2023, defines Gilbert syndrome phenotypically: isolated, asymptomatic unconjugated hyperbilirubinaemia, usually between 17 and 70 µmol/L, with normal ALT, aspartate aminotransferase (AST), alkaline phosphatase and gamma-glutamyltransferase (GGT), reticulocytes below 1.5 % and a negative Coombs test; hepatic ultrasound is performed only in selected cases. The patient is reassessed after 6 to 12 months and, if no new symptoms or abnormalities emerge, the diagnosis is established [9]. From this perspective, UGT1A1 genotyping is unnecessary and should not be used to confirm or redefine the diagnosis. It is reserved for distinguishing Gilbert syndrome from type 2 when bilirubin is higher, for family counselling and before the use of medicines that interact with UGT1A1; provocation tests with nicotinic acid, phenobarbital, rifampicin or fasting are considered obsolete [9].

The 2025 Chinese consensus accepts a clinical diagnosis after two total bilirubin determinations above the upper limit, separated by more than six months, predominantly indirect and without raised liver enzymes, after excluding haemolytic disorders. Genotyping is presented as a useful confirmatory tool (recommendation 1, 98.2% agreement) [2]. Both approaches agree on excluding haemolysis and liver disease but differ on genetic testing. This also reflects allele distribution, because coding variants predominate in East Asia and cannot be inferred from phenotype alone [2, 9].

5.4. Crigler–Najjar syndrome diagnosis

The diagnosis of Crigler–Najjar syndrome rests on persistent, marked unconjugated hyperbilirubinaemia without haemolysis and with normal liver function, and is confirmed by UGT1A1 analysis, which obviates liver biopsy [2, 5]. The Chinese consensus accepts genotyping as a first-line test for suspected Crigler–Najjar syndrome and defines type 2 by total bilirubin persistently above 102.6 µmol/L and type 1 by values equal to or greater than 342.0 µmol/L [2]. Analysis should include the exons, flanking intronic regions and promoter with its TATA element [15]. In the Croatian series, amplification of the TATA region by polymerase chain reaction (PCR) showed a TA7/TA7 genotype in three patients and TA6/TA6 in two, which did not explain the need for daily phototherapy. Complete sequencing of the coding region and splice sites subsequently revealed homozygous type 1 variants in all seven patients studied [11]. This example shows that finding *28 does not conclude the investigation of severe hyperbilirubinaemia.

The response to phenobarbital serves as a functional test: in type 2, bilirubin falls by approximately 25 to 30 %, with an effect usually apparent within two to three weeks, whereas there is no response in type 1 [2, 15]. In a 2026 Portuguese case, the team did not perform the challenge because bilirubin remained below 20 mg/dL and type 1 was unlikely [31]. The challenge entails exposure to a sedative medicine, should be interpreted in the light of age and spontaneous variability in bilirubin, and does not replace genotyping. In the worldwide registry, combining phenobarbital with phototherapy provided no benefit in patients without residual activity [12]. A lack of response may also result from variants that prevent induction, such as a three-nucleotide insertion in the HNF-1α binding site described in one patient [15].

In acute encephalopathy, brain MRI may show symmetrical T1 hyperintensity in the globus pallidus, thalami and basal ganglia; in chronic disease, bilateral T2 hyperintensity may occur in the same regions [2]. Liver biopsy is not recommended for diagnosis and is justified only when coexisting liver disease is suspected, an increasingly relevant possibility given fibrosis in patients with type 1 and p.(His39Asp) [2, 10, 12].

5.5. Practical algorithm

5.6. Genetic test limitations and report content

UGT1A1 sequencing detects point variants and small insertions or deletions, but counting the TA repeats in the promoter requires a method suited to the repeat, and larger deletions may escape methods based solely on sequencing; the Chinese consensus includes deletions among the causal changes in type 1 [2, 3]. The sources consulted provide no validated estimate of the clinical sensitivity of UGT1A1 analysis for each type. The worldwide registry states only that 83.7 % of the 221 patients had a genetic diagnosis, a proportion that measures the availability and outcome of testing in that cohort rather than analytical sensitivity [12]. A negative result does not exclude the diagnosis in the presence of an unequivocal phenotype.

The report should state the gene, versioned transcript (NM_000463.3, MANE Select), HGVS variant nomenclature and, for repeats, the TA unit count and historical designation; zygosity and, where two variants are present, phase in cis or trans established through parental studies; classification according to American College of Medical Genetics and Genomics and Association for Molecular Pathology criteria; and method limitations [4, 7, 32]. Phase is decisive in type 2: a coding variant combined with *28 on the other allele produces a different phenotype from the same combination in cis. In the Portuguese case, inability to test the parents left inheritance undetermined [1, 31]. ClinVar gives conflicting classifications and a “drug response” label to *28 and *6, requiring clinical interpretation rather than a binary reading of “pathogenic” or “benign” [32]. A compatible bilirubin phenotype or phenobarbital response does not make an uncertain variant pathogenic; family segregation and functional studies should be described as further investigation [7].

6. Differential Diagnosis

The differential diagnosis is organised according to the predominant bilirubin fraction and the presence of haemolysis or liver injury. The first table brings together the causes of predominantly unconjugated hyperbilirubinaemia; the second lists genetic causes of conjugated hyperbilirubinaemia and cholestasis that may present in newborns and infants; and the third compares the five inherited hyperbilirubinaemias without liver injury.

6.1. Predominantly unconjugated hyperbilirubinaemia

Neonatal causes frequently overlap. Gilbert syndrome alone rarely explains severe neonatal hyperbilirubinaemia but increases risk when combined with G6PD deficiency, Coombs-negative ABO disease, spherocytosis or prolonged breastfeeding [1, 39]. In Crigler–Najjar type 1, an additional haemolytic cause may precipitate crises, so G6PD and erythrocyte morphology deserve investigation even after the UGT1A1 genotype is identified [2, 14]. Congenital erythropoietic porphyria is relevant for another reason: affected newborns should avoid phototherapy for hyperbilirubinaemia to prevent severe photosensitivity [50].

6.2. Genetic conjugated hyperbilirubinaemia and infantile cholestasis causes

Cholestasis affects approximately 1 in 2,500 term newborns; biliary atresia accounts for 25–40 % of cases and monogenic disorders for approximately 25 % [44, 45]. Investigation includes GGT (low or normal in progressive familial intrahepatic cholestasis types 1 and 2, ARC syndrome and bile acid synthesis defects, and elevated in most other conditions), cholesterol, urinary reducing substances, thyroid and pituitary function, screening for congenital infections, ultrasound, hepatobiliary scintigraphy or biopsy, bile acids, alpha-1 antitrypsin, very-long-chain fatty acids, 7-dehydrocholesterol, spinal radiography, echocardiography, ophthalmological examination and, where appropriate, multigene panels, exome or genome sequencing [1, 44].

6.3. Inherited hyperbilirubinaemias without liver injury

Dubin–Johnson syndrome is more frequent than Rotor syndrome [7]. In the largest genetically confirmed neonatal series, 28 of 533 infants with cholestasis followed at a Saudi centre between 2008 and 2019 (5.3 %) had Dubin–Johnson syndrome. They were well-appearing term newborns without hepatosplenomegaly, with cholestasis and normal synthetic function from the first week, which resolved at 3–6 months; 43 % had recurrent episodes of jaundice over a median follow-up of 9.25 years, and the ABCC2 variant p.Gly758Val was found in 23 patients from 19 unrelated families [51]. This proportion reflects the caseload of a referral centre rather than population prevalence. In the same group, normal ALT in an infant with cholestasis had a sensitivity of 93 % and specificity of 90 %, but a positive predictive value of 34 %. A urinary coproporphyrin I proportion above 80 % had sensitivity, specificity and predictive values of 100 % in this sample of 28 cases and 20 controls [52]. Orphanet records an unknown prevalence in the general population and up to 1 in 1,300 among Jews of Iranian or Moroccan origin because of founder variants [34]. Rotor syndrome has a prevalence below 1 per million; insertion of a LINE-1 retrotransposon of approximately 6.1 kb into intron 5 of SLCO1B3 has a carrier frequency of 10.1 % in East Asian populations and 18.5 % in southern Han Chinese, and is not detected by routine sequencing [7].

7. Treatments and Risk Management Strategies

7.1. Principles by entity

The therapeutic strategy depends on residual UGT1A1 activity. Gilbert syndrome does not require treatment; intervention consists of recognition, explaining its benign nature and pharmacogenetic management [2, 9]. In type 2, the prognosis is generally favourable and phenobarbital is used when bilirubin is high, when aggravating factors are present or when jaundice compromises quality of life [2, 5]. In type 1, the aim is to prevent bilirubin encephalopathy through daily phototherapy from the neonatal period, treat decompensations and decide the timing of liver transplantation, which remains the only curative treatment [2, 5, 12]. Dubin–Johnson and Rotor syndromes do not require treatment [2, 7]. No medicine is specifically authorised for Crigler–Najjar syndrome in the European Union, so the medicines mentioned here for this indication are used off-label [15, 19].

Newborns with hyperbilirubinaemia of as yet unknown cause are treated according to general neonatal guidelines. The AAP recommends intensive phototherapy at thresholds defined by gestational age, age in hours and neurotoxicity risk factors (key action statement 10), escalation of care when bilirubin reaches 2 mg/dL below the exchange transfusion threshold, with intravenous hydration and emergency intensive phototherapy (statements 17 and 19), and urgent exchange transfusion in the presence of signs of intermediate or advanced encephalopathy or bilirubin at the corresponding threshold (statements 22 and 23) [14]. The bilirubin (mg/dL)/albumin (g/dL) ratio may complement the decision: ≥8.0 at 38 or more weeks without risk factors, ≥7.2 at 38 or more weeks with one risk factor or between 35 and 37 weeks without risk factors, and ≥6.8 between 35 and 37 weeks with at least one risk factor [14]. NICE recommends double-volume exchange transfusion without interrupting intensified phototherapy and without albumin priming [13]. The two guidelines differ regarding intravenous immunoglobulin for isoimmune haemolytic disease: the AAP presents 0.5 to 1 g/kg over 2 hours as an option, repeatable after 12 hours, while NICE specifies 500 mg/kg over 4 hours when bilirubin continues to rise by more than 8.5 µmol/L per hour despite phototherapy [13, 14]. Neither guideline defines a maximum total dose or a weight limit for immunoglobulin in the newborn [13, 14]. NICE advises against the use of barbiturates, metalloporphyrins, albumin, charcoal, cholestyramine, agar, clofibrate, D-penicillamine, glycerine, manna, riboflavin, traditional Chinese medicine, acupuncture and homoeopathy in neonatal jaundice in general [13]. This recommendation applies to neonatal jaundice without a specific diagnosis and does not contradict phenobarbital use in established type 2 or albumin use for type 1 decompensations at specialist centres, which are supported by patient series and consensus statements [2, 16].

7.2. Long-term type 1 phototherapy

Phototherapy is the mainstay of treatment and should begin as soon as pathological hyperbilirubinaemia is recognised; in the Amish and Mennonite cohort, initiation after day 13 was associated with a 3.5-fold higher risk of kernicterus [16]. The worldwide registry defined sufficient phototherapy as daily exposure for 6 to 12 hours to special blue fluorescent tubes or high-intensity light-emitting diodes (LED), with emission at 400 to 525 nm and a peak between 450 and 460 nm, skin irradiance of at least 40 µW/cm²/nm and up to 100 or more, exposure of 35 to 50 % of the body surface and a distance of 30 to 60 cm, with the aim of maintaining total bilirubin at 340 µmol/L or below. It also recommends white sheets or reflective surfaces, regular replacement of tubes (from 4 weeks to 4 months) and, when bilirubin rises, adding a daytime period rather than extending the night-time period, because efficacy decreases during the session [12].

In the North American cohort, light was administered for 9.2 ± 1.1 hours per day (range 7.0–12.5), with an irradiance of 33 to 103 µW/cm²/nm, maintaining a mean bilirubin concentration of 17.0 ± 5.9 mg/dL and a bilirubin/albumin ratio of 0.47 ± 0.17. LEDs proved 44 to 109 % more powerful than the reference fluorescent tubes, and low-irradiance fibre-optic systems are unsuitable for severe hyperbilirubinaemia [16]. Despite adherence, total bilirubin and the bilirubin/albumin ratio increased by an average of 0.46 mg/dL and 0.013 per year, reaching values that saturate 75 to 95 % of intravascular binding capacity at 18 years [16]. In the worldwide registry, the 57 patients receiving high-quality phototherapy throughout life (9.3 ± 2.3 hours per day) maintained bilirubin concentrations of 298.3 ± 90.3 µmol/L, and 92.9 % survived, most to adulthood without encephalopathy [12]. Equipment providing sufficient irradiance generates heat, so many patients require cooling at night. Eye protection remains recommended with LED sources, although the absence of ultraviolet radiation from current sources makes an increased risk of skin cancer unlikely [12, 16].

Effectiveness decreases with age because skin thickness and pigmentation increase and the surface area-to-weight ratio falls. Adherence is the principal limitation, as illustrated by the Croatian series [2, 11]. Reported adverse effects of prolonged phototherapy include xerosis, lichenification, hyperpigmentation and skin eruptions; one patient in the worldwide registry developed melanoma, without an established causal relationship [2, 12].

7.3. Adjunctive measures and type 1 decompensations

Decompensations are usually triggered by intercurrent febrile illnesses, fasting or failures of adherence [11]. In the North American centre’s protocol, inpatient management includes continuous phototherapy for approximately 24 hours per day, saline at 10–20 mL/kg, dextrose at 1 to 1.5 times maintenance requirements and avoidance of medicines that displace bilirubin from albumin; intravenous albumin is planned for a bilirubin/albumin ratio equal to or greater than 0.7, and exchange transfusion is considered from 0.9 [16]. The albumin dose stated in the article text is 1 to 2 g/kg per dose, as often as every 6 hours, and each infusion reduces the bilirubin/albumin ratio by 10 to 20 %, followed by a transient rise in bilirubin because of redistribution of pigment bound to tissues; the table in the same article prints “mg/kg”, which is an evident typographical error, so the text is followed [16]. The recommended dose is that of the protocol; for comparison, the United States label for 25% human albumin (Plasbumin-25®) provides, in haemolytic disease of the newborn, 1 g/kg approximately 1 hour before exchange transfusion to bind free bilirubin, with caution in hypervolaemic infants; the protocol’s dose of 1 to 2 g/kg as often as every 6 hours exceeds this posology, applied here by analogy, and lies outside the label conditions [55]. The protocol does not define a maximum cumulative dose; for hypovolaemia, the label limits the total dose to approximately 2 g/kg body weight in the absence of active bleeding, and safety and efficacy have not been established in the paediatric population [55]. All these measures are used off-label and derive from a single-centre series.

Plasma exchange rapidly reduces bilirubin during crises and has also been used during pregnancy to protect the fetal nervous system; it requires several sessions and large volumes of plasma and carries a risk of allergic reactions [2]. In a 15-year-old adolescent with type 2 and bilirubin-induced neurological dysfunction, seven daily sessions of therapeutic plasma exchange using fresh frozen plasma as the replacement fluid reduced total bilirubin from 35.0 to approximately 18.8 mg/dL, with improvement in alertness and abnormal movements and no pattern of acute encephalopathy on subsequent magnetic resonance imaging; this is an isolated report, which the authors themselves present as a rescue strategy rather than established practice [38]. For gallstones and biliary sludge, the North American centre’s protocol uses ursodeoxycholic acid at 15 to 30 mg/kg per day and enteral calcium carbonate at 2.5 mmol/kg per day in children and 100 mmol per day in adults [16]. For comparison, the labels give other posologies: the Portuguese SmPC for Ursofalk® recommends 10 mg/kg/day for dissolution of cholesterol gallstones, 12 to 16 mg/kg/day in primary biliary cholangitis and, in children aged 6 to 18 years with cystic fibrosis, 20 mg/kg/day in 2 to 3 divided doses, which may be increased to 30 mg/kg/day; the FDA label for ursodiol specifies 8 to 10 mg/kg/day in 2 to 3 divided doses for gallstone dissolution [56, 57]. The protocol’s recommended dose of 15 to 30 mg/kg/day exceeds the labelled posology for gallstones and coincides only, at its upper limit, with that for cystic fibrosis, so it lies outside the label conditions; the protocol does not define a maximum daily dose in absolute terms, and the SmPC gives 30 mg/kg/day as the ceiling of dose escalation in cystic fibrosis [56]. The Portuguese SmPC for calcium carbonate (Calcioral® 1250 mg, equivalent to 500 mg or approximately 12.5 mmol of calcium) specifies 1 to 2 tablets per day in adults, approximately 12.5 to 25 mmol of calcium, and it is not indicated below 12 years of age; the protocol doses, used to bind bilirubin in the intestine, exceed the label and constitute use outside the label conditions [58]; in the worldwide registry, ursodeoxycholic acid was the most frequent concomitant medicine, used by 14.5 % of patients [12]. The review by Sambati and colleagues adds the recommendation of a light, lipid-rich meal at bedtime [15].

Among approaches that bind bilirubin in the intestine, orlistat was evaluated in a randomised, placebo-controlled, double-blind crossover trial in 16 patients, with 4 to 6 weeks per arm: it increased faecal fat excretion by 333 % and unconjugated bilirubin excretion by 43 %, and reduced the plasma concentration of unconjugated bilirubin by 9 %, with a clinically relevant fall (greater than 10 %, mean 21 %) in 7 of the 16 patients [59]. The review by Sambati and colleagues presents these results differently, attributing a reduction of approximately 43 % to 40 % of patients; the abstract of the original trial is followed [15, 59]. Fenofibrate at 250 mg per day for one month did not alter bilirubin in two patients with type 2, and evidence of the efficacy and safety of metalloporphyrins is insufficient, with risks of photosensitivity and anaemia [15].

7.4. Phenobarbital in type 2 and Gilbert syndrome

Phenobarbital increases UGT1A1 transcription and acts only when residual activity is present. It reduces type 2 bilirubin by approximately 25 to 30%, usually visibly within two to three weeks, but is ineffective in type 1 [2, 12]. In the worldwide registry, 15 patients receiving phenobarbital alone, mostly with less severe phenotypes, maintained bilirubin of 146 ± 77.6 µmol/L without neurological symptoms or deaths; adding phenobarbital to phototherapy without residual activity provided no benefit [12].

The two sources differ in how they express paediatric dosing — total daily dose versus dose per administration — and both are recorded without combining them. The optimal duration of treatment has not been established; the literature describes two weeks to one month in adults and two to three weeks to several months or years in children [2]. Phenobarbital has no approved indication for Crigler–Najjar syndrome, and its use in this disease is off-label. For comparison with the consensus doses, the Portuguese SmPC for Bialminal® (100 mg tablets) specifies, as an anticonvulsant, 50 to 300 mg/day in adults and 3 to 6 mg/kg/day in children and, as a sedative, 30 to 120 mg/day in adults and 2 mg/kg three times a day in children, with a maximum dose of 600 mg in 24 hours; the SmPC for Luminaletas® (15 mg tablets) specifies 1 to 6 mg/kg/day in paediatric epilepsy [60, 61]. In the United States, the only FDA-approved phenobarbital is the injectable Sezaby®, restricted to neonatal seizures [62]. The posologies of the Chinese consensus lie within these ranges of the Portuguese SmPC; the divergence concerns the indication, not the dose. In pregnancy, the Chinese consensus states that doses above 750–1,000 mg per day early in gestation have been associated with facial malformations and intellectual disability, recommends keeping the dose below 50–60 mg per day and monitoring maternal bilirubin with targets of total bilirubin below 200 µmol/L and a bilirubin/albumin molar ratio below 0.5, and warns about effects on the infant through breast milk, with consideration of withdrawal if excessive somnolence or poor weight gain occurs [2]. The suggested strategy is phototherapy early in pregnancy and phenobarbital in the later stages [2]. Phenobarbital may also accelerate the inactivation of other glucuronidated medicines through enzyme induction [2].

7.5. Liver transplantation and cell therapies

Liver transplantation restores UGT1A1 activity and normalises bilirubin; it is the only curative treatment for type 1 and should be performed before irreversible neurological injury develops [2, 5]. The Chinese consensus recommends considering it before indirect bilirubin exceeds 300 µmol/L despite phototherapy, or in the presence of irreversible brain injury (recommendation 4, 99.1 % consensus). It lists the following indications: an insufficient response to phototherapy; intolerance of phototherapy because of severe photosensitive dermatitis or retinal injury; early neurological signs even with controlled bilirubin; and sleep deprivation, psychological disturbances or social isolation caused by prolonged phototherapy [2].

Reported modalities include whole or partial orthotopic transplantation and auxiliary partial orthotopic transplantation, with similar graft and patient survival in the compared series. Auxiliary transplantation preserves native liver tissue, retaining the theoretical option of future therapies, but complications occurred in 40 to 83% in the literature reviewed by Strauss and colleagues [11, 16]. Sambati and colleagues state that transplantation is usually recommended after the first year, with poorer outcomes in recipients under one [15]. Mortality and complication figures differ between cohorts; direct comparisons between series with different populations, periods and modalities are invalid.

Hepatocyte transplantation remains investigational. The first published case, a 10-year-old weighing 37 kg who received 7.5 × 10⁹ cells, achieved approximately 50% bilirubin reduction with enzyme activity up to 5.5% of normal and underwent liver transplantation four years later [15, 64]. Fourteen patients with type 1 received hepatocytes worldwide, 71% as a bridge to liver transplantation performed 4 to 31 months later; the effect is transient, lasting approximately 6 to 9 months [15]. In the Croatian series, one patient did not respond and underwent liver transplantation six months later [11].

7.6. Gene therapy: current status

Liver-directed gene therapy is investigational, and no product is authorised for Crigler–Najjar syndrome [15, 19]. The most advanced programme is GNT0003, a serotype 8 adeno-associated viral vector encoding UGT1A1, administered as a single intravenous infusion.

The published dose-escalation study reported no serious adverse events. Headache and liver enzyme abnormalities were the most frequent events, with ALT above the upper limit in four patients, possibly reflecting an immune response to the vector, treated with a glucocorticoid course [17]. Interpretation has clear limitations: five patients, no comparison group, a biochemical outcome and follow-up of approximately 78 to 80 weeks. Pre-existing capsid immunity excludes some patients, while vector dilution in the growing liver limits paediatric use [15, 17].

Regarding regulatory status and recent developments, the levels of evidence should be distinguished. GNT0003 has held orphan medicinal product designation in the European Union since 15 October 2014 (EU/3/14/1338) and appears on the EMA list of medicines in the PRIME scheme, with eligibility granted on 26 January 2023, in the list dated 23 September 2026 [19, 20]. Other orphan designations for this indication include adult human liver stem cells (EU/3/07/506, 2007) and modified messenger RNA encoding UGT1A1 protein (EU/3/16/1684, 2016); two designations for AAV8 vectors were withdrawn at the sponsors’ request in 2021 (EU/3/14/1321 and EU/3/16/1772) [19]. In the United States, the FDA orphan drug designation database records four designations for Crigler–Najjar syndrome, none with approval for the orphan indication: heterologous human liver progenitor cells from Promethera Biosciences, designated on 9 March 2012; a non-replicating recombinant serotype 8 adeno-associated vector expressing isoform 1A1 of bilirubin-UDP-glucuronosyltransferase from Audentes Therapeutics, on 24 May 2016; modified messenger RNA encoding UGT1A1 protein, formulated in lipid nanoparticles, from Moderna Therapeutics, on 5 July 2016; and Généthon’s rAAV8-hUGT1A1, on 29 October 2024, with the indication worded as treatment of Crigler–Najjar syndrome in adults and children aged 10 or more years who require phototherapy for 6 or more hours per day [18]. The same database also contains a 1985 designation for flumecinol for the treatment of newborn hyperbilirubinaemia unresponsive to phototherapy, with a status of designation withdrawn or revoked and no approval [18]. Généthon’s United States designation is ten years later than the European designation for the same programme, and its wording matches the target population described in the PRIME scheme entry. In October 2025, a joint announcement by Généthon and Hansa Biopharma described the treatment of a female patient with anti-AAV antibodies, in whom imlifidase administered before gene therapy cleaved and inactivated the antibodies and enabled GNT0003 administration, with a fall in bilirubin and discontinuation of phototherapy 16 weeks later, as specified in the protocol, and no serious adverse effects attributed to the two medicines [71]. This concerns a single patient; the data were first released in a sponsor announcement and later described in an opinion article signed by the clinical team, Généthon and Hansa Biopharma in the May 2026 issue of Molecular Therapy, with restricted access, so the values given here are taken from the announcement [71]. By 7 October 2026, no publication of the pivotal-phase results or news of a marketing authorisation application to the EMA or FDA had been found in the sources consulted; this negative search does not prove that such developments do not exist. The continued inclusion of GNT0003 on the “current” PRIME list is consistent with the absence of a submitted application, according to the note accompanying that list [20].

7.7. Pharmacogenetic management: UGT1A1 as a toxicity determinant

Reduced glucuronidation capacity affects more than bilirubin. The same genotype that causes Gilbert syndrome reduces the clearance of medicines that depend on UGT1A1 and increases the risk of toxicity; conversely, medicines that inhibit the enzyme raise indirect bilirubin without liver injury [2, 9]. CPIC classifies genotypes as normal metabolisers (two reference-function or increased-function alleles, such as *1/*1, *1/*36, *36/*36), intermediate metabolisers (one decreased-function allele, such as *1/*28, *1/*37, *1/*6) and poor metabolisers (two decreased-function alleles, such as *28/*28, *28/*37, *37/*37, *6/*6) [21]. The *80 allele (rs887829, c.-364C>T) is in almost complete linkage disequilibrium with *28 and *37 and serves as a surrogate marker in some assays, although it does not itself influence gene expression [3, 21].

The relative activity estimated in the Medical Genetics Summaries chapter is 100 % for *1, 130 % for *36, 70 % for *6, 65 % for *28, 50 % for *27 (p.Pro229Gln) and 50 % for *37; one copy of *28 reduces transcriptional activity by approximately 35 % and two copies by approximately 70 % [3]. The *36 allele is associated with greater promoter activity and a lower risk of neonatal hyperbilirubinaemia [3].

The three irinotecan guidelines differ in the magnitude of dose reduction (70 % of the dose in the DPWG guideline, a 30 % reduction in the CERSI-PGx guideline and Italian recommendation, and a reduction of at least one dose level in the United States label) and in defining whom to test; the French guideline cited by the DPWG does not recommend genotyping for doses below 180 mg/m² [22–24]. These differences arise from different starting points regarding the reference dose rather than contradictory data. The editorial accompanying the British guideline notes that neither the United States nor the European label contains a specific recommendation for testing before treatment, and estimates that 79 patients need to be genotyped to prevent one episode of grade 3 or higher neutropenia and 127 to prevent one episode of grade 3 or higher diarrhoea, compared with 210 to 251 for DPYD testing, which is already widely implemented [82]. The body surface area posologies in these tables apply to adult patients, the only population in the approved indications of the labels consulted, and none of these labels sets an absolute maximum dose: the limits derive from the dose per cycle and from the criteria for dose reduction because of toxicity [72, 77].

The quantitative data should be interpreted with attention to the source population. In the sacituzumab govitecan label, among 1,202 patients with a known genotype, grade 3–4 neutropenia occurred in 57 % of *28/*28 patients (n = 138) and 48 % of *1/*28 patients (n = 531), and grade 3–4 anaemia occurred in 17 % and 9 %, respectively. The median time to the first episode of neutropenia was 9, 19 and 21 days in variant homozygotes, heterozygotes and wild-type homozygotes [74]. The European SmPC for the same medicine presents different percentages, obtained from another analysis (59.6 %, 50.7 % and 44.9 % for grade 3–4 neutropenia), and adds the *6 allele, which is absent from the United States label, stating that 27 % of the East Asian population is heterozygous for *6 [75]. The two sets of numbers should not be combined. Patients with known Gilbert syndrome were excluded from the pivotal trials of this medicine [74, 75].

For atazanavir, the positive predictive values for bilirubin-related discontinuation in carriers of rs887829 T/T were 60 % (95 % confidence interval 32–84) in white patients, 29 % (8–58) in Hispanic patients and 20 % (9–36) in black patients, in a trial with 481 participants. In a Swiss cohort, the estimated cumulative discontinuation rates during the first year were 63 %, 24 % and 15 % for two, one and no decreased-function alleles, respectively [21]. CPIC had no paediatric data and directly adapted the adult recommendations to children, which is an explicitly stated extrapolation rather than paediatric evidence [21]. In a 2025 annotation, the DPWG recommends avoiding atazanavir in *28/*28 patients if a suitable alternative is available and considers that no action is necessary for the other phenotypes, including poor metabolisers with other alleles, which differs from CPIC [83]. Newborns exposed to atazanavir in utero should be monitored for severe hyperbilirubinaemia during the first days of life; in the study cited by the European SmPC, six of 40 infants received phototherapy for a maximum of 4 days and there were no cases of kernicterus [78, 79].

The Chinese consensus summarises this area in recommendation 5, with 95.5 % agreement: perform UGT1A1 genotyping before starting irinotecan, atazanavir and belinostat to assess adverse-effect risk [2]. It adds that paracetamol, statins, oestradiol and lamotrigine are metabolised by UGT1A1, that sorafenib and lenvatinib inhibit the enzyme and may raise bilirubin, and that without other hepatic injury markers this elevation should be assessed and monitored rather than prompting unnecessary drug withdrawal [2]. Compounds competing with bilirubin or inhibiting UGT1A1 also include tyrosine kinase inhibitors such as regorafenib, sorafenib, lapatinib, nilotinib, sunitinib and pazopanib, non-steroidal anti-inflammatory medicines, ezetimibe, oxazepam, lorazepam, ethinylestradiol, ciclosporin A, rifampicin, buprenorphine, tocilizumab and natural compounds such as silymarin complex flavonolignans, quercetin and epigallocatechin gallate [9].

7.8. Regulatory comparison between the United States and European Union

Orphan designation in the United States or European Union and inclusion in the PRIME scheme are statuses that support development and do not constitute a favourable opinion or authorisation. The FDA database itself distinguishes designation status from approval status for the orphan indication, and none of the products for this disease has approval [18]. Authorisation, in turn, does not guarantee marketing or reimbursement in each country. No data on access in Portugal (INFARMED) or Brazil (ANVISA, CONITEC) to these products for Crigler–Najjar syndrome were found in the sources consulted, which does not demonstrate that such data do not exist.

7.9. Trial synthesis and interpretive limitations

These three sources are not directly comparable. The gene therapy study involved dose escalation in five patients, without a comparison group and with a biochemical outcome; it does not allow estimates of population efficacy or durability beyond the reported follow-up. The orlistat trial was randomised and used a crossover design, but had a small mean effect and a benefit concentrated in a minority of patients, illustrating the distinction between the group mean and the individual response. The Dutch study was observational and its data were read only in the abstract; it shows that dose reduction brings the risk in poor metabolisers closer to that in the remaining patients, rather than showing that dose reduction is unnecessary. In all cases, a relative reduction should be interpreted alongside the residual absolute rate, and the absence of a statistically significant difference in a study of this size does not establish equivalence.

8. Reproductive Options and Family Counselling

8.1. Crigler–Najjar syndrome

Crigler–Najjar syndrome types 1 and 2 are autosomal recessive: the parents of an affected child are generally heterozygous, and each subsequent pregnancy of the couple has a 25 % probability of an affected child, a 50 % probability of an asymptomatic carrier and a 25 % probability of a non-carrier [1, 5]. Carrier detection is possible in families in which the causal variant has been identified, and prenatal diagnosis is available when the diagnosis has molecular or biochemical confirmation [1]. The estimated carrier frequency for type 1 in the Western hemisphere is approximately 1 in 1,000, but is much higher in founder populations, such as the Mennonites and Amish of Pennsylvania, and in communities with high consanguinity, such as the Tunisian community [1, 12, 16]. In the worldwide registry, 43.3 % of patients were children of consanguineous unions [12].

Counselling must consider the spectrum associated with the same gene. Heterozygotes for a Crigler–Najjar-type variant may have Gilbert-type hyperbilirubinaemia when the structurally normal allele carries the A(TA)7TAA promoter, and intermediate bilirubin concentrations are common in relatives [1, 8]. A child may inherit a coding variant from one parent and a *28 allele from the other, with a phenotype intermediate between type 2 and Gilbert syndrome, as in the Portuguese case described in section 10 [31]. Parental studies are necessary to establish the phase of the variants and estimate recurrence risk correctly [1, 31].

Reproductive decisions also depend on the perception of severity and the family’s experience with phototherapy. In the Croatian series, the parents of one sibship declined prenatal diagnosis and all three children had type 1 [11]. When the familial variant is known, prenatal diagnosis is a technical option [1]; preimplantation genetic testing is also possible under these circumstances, as GeneReviews describes for hereditary Rotor hyperbilirubinaemia [7]. The discussion should include the daily burden of phototherapy, the prospect of transplantation and the continuing experimental status of gene therapy [12, 17].

8.2. Pregnancy in a woman with Crigler–Najjar and the neonatal period

Elevated maternal bilirubin may increase the risk of neonatal kernicterus [2]. The Chinese consensus proposes maintaining maternal total bilirubin below 200 µmol/L and the bilirubin/albumin molar ratio below 0.5; the review by Sambati and colleagues gives a target below 11.7 mg/dL with a ratio below 0.5 [2, 15]. In type 2, the strategies described include phenobarbital, phototherapy and albumin infusions, with phototherapy suggested early in pregnancy and phenobarbital in the later stages; plasma exchange may prevent transplacental passage of bilirubin in severe situations [2]. Phenobarbital doses during pregnancy and breastfeeding are discussed in section 7.4. The bilirubin of a newborn whose mother has Crigler–Najjar syndrome should be monitored, and the review by Sambati describes two children of mothers with type 1 who required exchange transfusion [15].

8.3. Gilbert syndrome

Gilbert syndrome alone does not justify prenatal diagnosis or reproductive decisions. Approximately 50 % of European and North American populations are heterozygous for the promoter insertion and 9 % are homozygous, so the condition frequently appears in successive generations without indicating dominant inheritance [1]. The central message for the family is the benign nature of the condition, the absence of a need for treatment and the importance of informing clinicians before receiving medicines that depend on UGT1A1 [2, 9]. Health education reduces anxiety and unnecessary investigations and includes avoiding fasting, dehydration and infections as triggers of jaundice [2]. A newborn with a sibling who required phototherapy has an increased risk of significant hyperbilirubinaemia, and this information contributes to neonatal assessment regardless of genotype [13, 14].

8.4. Dubin–Johnson and Rotor syndromes

Dubin–Johnson syndrome is autosomal recessive and benign [34]. In Rotor syndrome, the parents of an affected individual are presumed to be heterozygous for variants in SLCO1B1 and SLCO1B3 on the same chromosome; each sibling has a 25 % probability of being affected, and carriers of one, two or three pathogenic variants are asymptomatic and are not at risk of disease [7]. Because most patients with Rotor syndrome are born to consanguineous parents, the diagnosis may reveal consanguinity and warrant a genetics consultation; requests for prenatal diagnosis of a benign condition are rare, and the decision is personal [7].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Crigler–Najjar: global rarity and founder populations

Crigler–Najjar syndrome type 1 is rare in the Western hemisphere, with a prevalence below 1 per million and a carrier frequency of approximately 1 in 1,000, according to the Oxford Desk Reference [1]. The review by Sambati and colleagues estimates 0.6 to 1 case per million newborns worldwide and fewer than 1 per 100,000 people in Europe, and the European Medicines Agency (EMA), in its 2014 orphan designation, considered that the disease affected fewer than 0.1 in 10,000 people in the European Union [15, 19]. The heading of the Orphanet ORPHA:205 entry gives the class 1–9 per 100,000, but the text of the same entry states fewer than 1 per 100,000 in Europe; the text, which is consistent with the other sources, is followed [5]. These estimates combine both types and should not be applied to type 2 in isolation.

The frequency rises in populations with founder effects or consanguinity. Among Mennonites of the Weaverland and Groffdal conferences, the variant c.222C>A p.(Tyr74Ter) has an allele frequency of 2.17 % and an incidence of 1 in 2,124 births; among the Lancaster Old Order Amish, the frequency is 0.83 % and the incidence is 1 in 14,520, with the variant having been introduced separately into the two populations approximately 300 years ago [16]. In Tunisia, prevalence reaches 8 patients per million inhabitants; the variant c.1070A>G p.(Gln357Arg) arose approximately 32 generations ago and has also been identified in Bedouin families from Kuwait with the same ancestral haplotype, and consanguineous unions range from 25 % in central Tunisia to 60 % in rural areas [12, 33]. This variant accounted for 39.9 % of alleles in the worldwide registry, followed by p.(Tyr74Ter) (7.1 %), c.1220del (4.9 %), c.1489del (3.4 %) and p.(Trp461Arg) (2.2 %); c.1220del is concentrated in the southwestern Netherlands [12]. These proportions describe the centres contributing to the registry — 43.4 % of patients from North Africa, 38.5 % from Europe, 9.5 % from North America and 8.6 % from the Middle East and Asia — rather than the worldwide distribution of alleles [12].

In Croatia, eight patients with type 1 born between 1992 and 2022, during a period with 1,314,533 live births, correspond to an incidence of 6.1 per million; four descended from families originating in Janjevo, a Croatian enclave in Kosovo with prolonged reproductive isolation [11]. Because the eight patients belong to five families, two of which account for five patients, the authors acknowledge that familial clustering contributes to the estimate [11]. HGVS normalisation (section 2.2) indicates that the two AG deletions described in that series correspond to the same variant, which should be considered in future analyses of founder effects [4, 11]. In China, data from two hospitals specialising in liver disease in Beijing, between 2002 and 2023, showed 1.077 cases of Crigler–Najjar syndrome per 100,000 inpatients, a hospital measure rather than a population measure [2]. In Iran, a centre in Shiraz transplanted 69 patients between 2000 and 2023 [63].

9.2. Gilbert syndrome and UGT1A1 alleles by ancestry

The frequency of Gilbert syndrome depends on the definition of normal bilirubin, sex and ancestry. The Oxford Desk Reference gives 4–7 % of the general population, GeneReviews approximately 5–10 % of Europeans, and the review by Vítek and Tiribelli approximately 10 %, with a range of 2–12 % and a male predominance of 3:1 [1, 7, 9]. In a Czech sample of almost 2,600 people, prevalence using the threshold of 17 µmol/L was 8.9 % (11.6 % in men and 6.1 % in women), and the 95 % reference interval was 5.5–27.5 µmol/L in men and 3.8–19.9 µmol/L in women, leading the authors to advocate different thresholds by sex [9]. The Chinese consensus cites prevalences of 15–25 % in African populations, 2–10 % in Caucasian populations, up to 20 % in South Asia and the Middle East, and 3.2 % in Singapore, and notes the absence of large-scale epidemiological studies in China [2].

Bosma and colleagues found the A(TA)7TAA allele in 40 % of the alleles of 55 Dutch controls, and the Oxford Desk Reference estimates that approximately 50 % of European and North American populations are heterozygous and 9 % are homozygous [1, 8]. These are genotype frequencies: not all homozygotes have jaundice, and the figures should not be interpreted as the prevalence of the syndrome [3, 8]. In the Chinese population, the most common variants associated with Gilbert syndrome are c.-3275T>G (the consensus designation), A(TA)7TAA and c.211G>A [2].

9.3. Other differential entities

G6PD deficiency affects 13 % of African American men and approximately 4 % of African American women and is more common in populations from Africa, the Middle East, the Mediterranean, the Arabian Peninsula and Southeast Asia [14]. Dubin–Johnson syndrome has an unknown prevalence in the general population and reaches up to 1 in 1,300 among Jews of Iranian or Moroccan origin, according to Orphanet; the Chinese consensus attributes this frequency to Hispanic and Iranian Jewish populations, a difference that could not be resolved in the sources consulted, so the Orphanet entry revised in 2026 is followed [2, 34]. In Arab populations, the ABCC2 variant p.Gly758Val has a founder effect [51]. The LINE-1 insertion in SLCO1B3 responsible for Rotor syndrome is almost exclusive to East Asia [7].

9.4. Portugal and Brazil

In the sources consulted, Portuguese data are limited to a 2026 case report from Unidade Local de Saúde Arco Ribeirinho (Barreiro), concerning a newborn girl of Nepalese ancestry with a phenotype intermediate between Gilbert syndrome and Crigler–Najjar syndrome type 2 (section 10.2) [31]. No epidemiological data on Crigler–Najjar syndrome or UGT1A1 alleles in Portugal or Brazil were found in the underlying sources, which does not demonstrate that such data do not exist in the literature.

9.5. Inequality in treatment access

The most marked difference between populations concerns healthcare rather than genetics. In the worldwide registry, 75 patients who required phototherapy but lacked access to equipment of sufficient quality maintained total bilirubin concentrations of 443.6 ± 141.4 µmol/L; 42.7 % had neurological symptoms and 62.7 % died, at a median age of 4 years [12]. The authors attribute these differences to factors such as poverty, irregular access to electricity, a lack of home phototherapy systems and limited transplantation infrastructure [12]. In the compilation by Strauss and colleagues, brain injury occurred in 18–26 % of patients in developed countries and 50–93 % in countries with limited resources [16].

9.6. Reference networks and patient associations

The relevant European reference network is MetabERN, which includes the centre at Hôpital Antoine-Béclère (Assistance Publique–Hôpitaux de Paris), with a dedicated Crigler–Najjar syndrome programme, whose expert reviewed the Orphanet entry for the disease [5, 25]. ERN RARE-LIVER covers genetic cholestatic diseases, biliary atresia and alpha-1 antitrypsin deficiency, which are relevant to the differential diagnosis [26]. The European CureCN consortium, funded by the Horizon 2020 programme and coordinated by Généthon, brings together researchers and French, Italian and Dutch patient associations [27]. The worldwide registry acknowledges the Dutch Najjar Foundation, the French association for patients with Crigler–Najjar syndrome and the Italian association CI AMI [12]. In the United States, the Crigler-Najjar Association/King’s Way Foundation appears in the National Organization for Rare Disorders directory [85]. The Oxford Desk Reference lists the Children’s Liver Disease Foundation in the United Kingdom for childhood liver diseases [1].

10. Illustrative Clinical Case

10.1. Type 1 in a Croatian sibship: diagnosis, home phototherapy and adherence

In 2023, Kovačić Perica and colleagues described the eight patients with type 1 treated at University Hospital Centre Zagreb between 1992 and 2022 [11]. Family B illustrates the diagnostic and therapeutic course. Three siblings — two girls and one boy, patients 6, 7 and 8 in the series — were the only children of a mother with confirmed Gilbert syndrome (TA7/TA7 genotype); the family originated from the Croatian enclave of Janjevo in Kosovo, without known consanguinity. All were born at term after uncomplicated pregnancies, with normal birth weights, and developed progressive jaundice during the first days of life, with drowsiness and feeding difficulties, but normal stool and urine colour; examination showed neither dysmorphism nor hepatosplenomegaly. The maximum unconjugated bilirubin before phototherapy ranged from 374 to 521 µmol/L in the series, without haemolysis and with normal liver enzymes, albumin and coagulation [11].

The first test performed was analysis of the TATA region of the promoter. In the two siblings with recorded results, the genotype was TA6/TA6, which excluded Gilbert syndrome as the explanation; sequencing of the coding region showed homozygosity for c.1021C>T p.(Arg341Ter), designated *10, in exon 3, a variant previously described in Pakistani and Chinese patients [11]. The phenobarbital challenge, performed in all patients in the series, did not significantly alter bilirubin. The parents declined prenatal diagnosis in subsequent pregnancies [11].

Phototherapy was started during the first weeks and continued at home with dedicated equipment — an infant bed with light-emitting diode (LED) lamps providing a maximum intensity of 180 µW/cm²/nm and portable lamps providing 100 to 160 µW/cm²/nm. In hospital, the team confirmed that 10 hours of phototherapy per day, consistently administered, maintained bilirubin below neurotoxic values (<340 µmol/L). During the six months before publication, however, control was inadequate because of reduced adherence and frequent febrile illnesses [11]. Figure 1 of the article (PMC10620791, Figure 1) shows total bilirubin in the four patients receiving phototherapy: in patient 6, after values near 250 µmol/L during the first year, bilirubin rises again to approximately 460 µmol/L near 4.5 years; patient 7 has a peak near 500 µmol/L shortly after 2 years; patient 8, an infant, remains more stable. The authors attributed the upward trend lines in the older patients to poor adherence, whereas patient 5, with strict adherence and 9–11 hours of nightly phototherapy, maintained values of 180–230 µmol/L [11]. At that time, all three siblings had normal psychomotor development.

The contrast with patient 1 in the same series shows what is at stake. This boy discontinued phototherapy at 9 months and remained untreated for years; at 5 years, he developed abrupt regression of development and language, postural instability and choreoathetosis. He underwent an unsuccessful hepatocyte transplant and, six months later, a segmental orthotopic liver transplant; at 31 years, he has severe psychomotor delay and depends on others [11]. Of the four patients who received a transplant from a living related donor, three had successful outcomes, with normal liver function 11, 13 and 21 years later, and one female patient died four days after transplantation because of primary graft non-function [11].

This case does not establish routine management, but summarises several points of this article: a Gilbert genotype in the mother or patient does not explain bilirubin above 340 µmol/L; phenobarbital does not alter type 1; effective phototherapy depends on irradiance, duration and adherence; and intercurrent illnesses destabilise control. The decision about transplantation was made when bilirubin rose to potentially neurotoxic values despite phototherapy, which is consistent with the Chinese consensus recommendation to consider transplantation before indirect bilirubin exceeds 300 µmol/L or irreversible brain injury develops [2, 11].

10.2. Intermediate type 2/Gilbert phenotype in a Portuguese case

In 2026, Jesus Sá and colleagues, from Unidade Local de Saúde Arco Ribeirinho, published the case of a newborn girl born to non-consanguineous parents of Nepalese ancestry, at 37 weeks and 6 days, weighing 2,555 g [31]. Jaundice appeared within the first 24 hours (maximum transcutaneous bilirubin 12.3 mg/dL); the mother was O Rh-negative and the child O Rh-positive, with a negative direct Coombs test. She received two courses of phototherapy during the first days, with a loss of 9.6 % of her body weight and formula supplementation. At 12 days, she returned to the emergency department with marked jaundice extending to the palms and soles, total bilirubin of 29.7 mg/dL and direct bilirubin of 0.6 mg/dL; the full blood count, reticulocytes (1.5 %), blood film, coagulation, transaminases, GGT, alkaline phosphatase, albumin, urine, abdominal ultrasound and transfontanellar ultrasound were normal. After three days of phototherapy, bilirubin fell to 13.3 mg/dL [31].

Over the next two months, bilirubin ranged from 10.9 to 19.04 mg/dL without reaching the phototherapy threshold for her age. The team decided not to perform the phenobarbital challenge because they considered type 1 unlikely. UGT1A1 analysis revealed NM_000463.3:c.1456T>G p.(Tyr486Asp), associated with type 2 in homozygous form, and the promoter allele *28; the parents could not be studied, so the phase and mode of transmission remained undetermined. The child was breastfed until 12 months, with normal growth and neurodevelopment, and bilirubin normalised at 11 months [31]. Figure 1 of the article (PMC12875353, Figure 1) shows total bilirubin of 15.3 mg/dL at 21 and 26 days, 19.0 mg/dL at 1 month, 10.9 at 2 months, 8.5 at 4 months and 0.7 mg/dL at 11 months, with direct bilirubin always between 0.0 and 0.6 mg/dL.

The authors interpret the presentation as an intermediate phenotype — initial type 2 concentrations followed by normalisation compatible with Gilbert syndrome — and suggest that neonatal hepatic immaturity and breast milk jaundice amplified the deficiency [31]. This interpretation should be read with the limitation of unknown phase: definitive classification requires knowing whether the two changes are on different alleles. The case also shows why genotyping alone, without longitudinal follow-up, may assign an incorrect prognostic label to a newborn.

11. Prognosis and Long-Term Follow-up

11.1. Prognosis by entity

Gilbert syndrome has a very good prognosis, and drug interactions are the main health concern [9]. Crigler–Najjar syndrome type 2 generally has a favourable course, with survival to adulthood without neurological injury, but is not free from risk: cases of bilirubin encephalopathy in adults during acute insults have been reported, and a proportion of patients develop gallstones, fibrosis or cirrhosis [2, 21]. The German series of patients homozygous for p.(His39Asp) showed that half developed end-stage liver disease despite a mild bilirubin phenotype [10]. Type 1 is fatal without treatment: in the original description, six of seven infants died of bilirubin encephalopathy by 15 months, and in the worldwide registry the subgroup without adequate access to treatment, which represents natural history, had a mortality of 62.7 % and a median age at death of 4 years [12, 16]. With high-quality phototherapy maintained throughout life, survival exceeded 92 % and several patients reached their fourth decade without encephalopathy; with transplantation, bilirubin normalises, and graft and patient survival was 100 % in the North American cohort over 0.1 to 14.9 years [12, 16]. Dubin–Johnson and Rotor syndromes are benign, without cirrhosis or liver failure [7, 34].

11.2. Risks persisting during treatment

Biochemical control does not eliminate three problems. The first is the progressive loss of phototherapy efficacy: bilirubin and the bilirubin/albumin ratio rise by an average of 0.46 mg/dL and 0.013 per year, and by 18 years approach saturation of albumin-binding capacity [16]. The second is liver disease: advanced fibrosis was present in 16 % of the patients assessed in the worldwide registry and in 33 % of those dependent on phototherapy aged 7 or more years, 60 % of the livers examined in the North American cohort had mild to severe fibrosis, and half of the patients with p.(His39Asp) developed end-stage liver disease [10, 12, 16]. The third is gallstones, present in 15.4 % of ultrasound examinations in the worldwide registry and in 43 % of the North American cohort [12, 16]. Added to these is the impact of daily phototherapy on quality of life and adherence, recognised by patients and families as the greatest burden [11, 12].

11.3. Proposed follow-up

No surveillance guideline specific to Crigler–Najjar syndrome is available in the sources consulted, and GeneReviews states that no clinical practice guideline has been published for Rotor syndrome because neither treatment nor surveillance is recommended [7]. The following proposal brings together the elements supported by the sources, identifying the origin of each; the intervals do not appear as such in any guideline and should be adapted by the treating team.

In Gilbert syndrome, follow-up consists of confirming the stability of the presentation: if no symptoms or new haematological or biochemical abnormalities emerge after 6 to 12 months, the diagnosis is established and, according to the clinical experience of the review authors, rarely needs to be reconsidered [9]. No restrictions on diet or physical activity need to be recommended; indeed, the syndrome is more prevalent among elite athletes [9]. The clinically useful information to convey is the need to identify the condition before prescribing medicines that depend on UGT1A1, and to avoid interpreting hyperbilirubinaemia induced by these medicines as liver injury [2, 9].

12. Summary and Final Considerations

The genetic forms of neonatal jaundice with unconjugated hyperbilirubinaemia centre on one gene, UGT1A1, and a continuous spectrum of enzyme deficiency: Gilbert syndrome with approximately 30 % activity, Crigler–Najjar syndrome type 2 with less than 10 %, and type 1 with no detectable activity. The practical distinction rests on bilirubin concentration, persistence, response to phenobarbital and genotype, but the biochemical boundaries overlap and the same patient may change classification during the first year of life, as the Portuguese case described illustrates. The clinical consequence of this continuity is that the diagnostic label should not replace longitudinal monitoring.

Type 1 requires urgent action. Bilirubin rises by 4.3 mg/dL per day during the first days and reaches potentially neurotoxic concentrations between 5 and 14 days; delayed initiation of phototherapy increased the risk of kernicterus by a factor of 3.5 in a cohort with a founder variant. Well-conducted phototherapy — an appropriate source, measured irradiance, sufficient exposed surface and approximately 9 to 12 hours per day — enabled survival of more than 92 % of patients with access to treatment, many to adulthood without neurological injury. The main determinant of prognosis worldwide is not genetic: among patients without access to adequate equipment, 42.7 % had neurological symptoms and 62.7 % died, with a median age at death of 4 years.

Recent literature has revised three classical concepts that should be corrected in teaching. The first is that the liver is structurally normal in Crigler–Najjar syndrome: advanced fibrosis occurs in a substantial proportion of patients dependent on phototherapy, and a German series showed progressive liver disease in patients with a mild bilirubin phenotype. The second is that Gilbert syndrome is irrelevant: it is indeed benign, but modifies neonatal jaundice when it coexists with haemolysis and influences the toxicity of several medicines. The third is that elevated bilirubin during treatment with irinotecan, atazanavir or tyrosine kinase inhibitors indicates liver injury: in the absence of other markers, it reflects slower biotransformation and does not in itself justify discontinuing treatment.

The geneticist’s work in prolonged neonatal jaundice begins by distinguishing the conjugated from the unconjugated fraction, excluding haemolysis and reviewing newborn screening before directing molecular investigation. UGT1A1 analysis should cover the exons, intronic boundaries and promoter, and counting the TA repeats requires an appropriate method; finding the *28 allele does not conclude the investigation of severe hyperbilirubinaemia, as the Croatian series showed. The report should state transcript NM_000463.3 with its version, HGVS nomenclature, the number of repeats with the historical designation, zygosity, phase when two variants are present and the limitations of the method. Notations for the TA repeat differ between official sources and should be confirmed in reference databases, never converted manually.

In treatment, phototherapy and liver transplantation remain the two interventions with demonstrated results in type 1, and phenobarbital is useful only where inducible enzyme is present. Liver-directed gene therapy showed, in five patients, maintenance of bilirubin below the defined efficacy threshold without phototherapy for at least 78 weeks at the higher dose; it remains investigational, without authorisation in any of the jurisdictions consulted, and the pivotal-phase results had not been published by November 2026. UGT1A1 pharmacogenetics, by contrast, is already recommended practice for irinotecan by several societies, with actual differences in the magnitude of dose reduction, and for atazanavir by a CPIC guideline.

Open questions remain and should be stated as such: the cause of liver fibrosis in these patients, the ideal timing of transplantation, the durability of gene therapy and management of patients with pre-existing immunity to the vector, the clinical sensitivity of molecular analysis by disease type, and the absence of Portuguese or Brazilian epidemiological data on these conditions in the sources consulted. None of these gaps justifies delaying the essentials: in a newborn with persistent jaundice, measuring and fractionating bilirubin, treating without awaiting the molecular diagnosis and ensuring continued access to phototherapy are what distinguish normal development from kernicterus in the published series.

13. Bibliographic References

1. Firth HV, Hurst JA. Prolonged neonatal jaundice and jaundice in infants below 6 months. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 288–91.
2. Zheng S, Xu X, Nan Y, Hou W, Bai J, Tang S, et al.; Inherited Metabolic Liver Disease Collaboration Group, Chinese Society of Hepatology, Chinese Medical Association. Expert Consensus on the Diagnosis and Management of Inherited Hyperbilirubinemia (2025). J Clin Transl Hepatol. 2026;14(1):83–95. doi:10.14218/JCTH.2025.00440. PMID:41659992.
3. Dean L, Kane M. Atazanavir Therapy and UGT1A1 Genotype. In: Pratt VM, Scott SA, Pirmohamed M, Esquivel B, Kattman BL, Malheiro AJ, editors. Medical Genetics Summaries [Internet]. Bethesda: National Center for Biotechnology Information; 2012–. Created 17 October 2023. NBK596252. Accessed 7 October 2026.
4. VariantValidator (version 4.0.1, hgvs 4.0.0). Normalisation of variants in UGT1A1 on MANE Select transcript NM_000463.3. [https://variantvalidator.org](https://variantvalidator.org). Accessed 7 October 2026.
5. Orphanet. Crigler-Najjar syndrome. ORPHA:205. Reviewer: Labrune P (MetabERN). Last updated June 2021. [https://www.orpha.net/en/disease/detail/205](https://www.orpha.net/en/disease/detail/205). Accessed 7 October 2026.
6. Orphanet. NON RARE IN EUROPE: Gilbert syndrome. ORPHA:357. Last updated 1 July 2026. [https://www.orpha.net/en/disease/detail/357](https://www.orpha.net/en/disease/detail/357). Accessed 7 October 2026.
7. Jirsa M, Knisely AS, Schinkel A, Kmoch S. Rotor Syndrome. In: GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published 13 December 2012; comprehensively updated 27 February 2025. NBK114805. PMID:23236639.
8. Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A, Oostra BA, et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyltransferase 1 in Gilbert's syndrome. N Engl J Med. 1995;333(18):1171–5. doi:10.1056/NEJM199511023331802. PMID:7565971.
9. Vítek L, Tiribelli C. Gilbert's syndrome revisited. J Hepatol. 2023;79(4):1049–55. doi:10.1016/j.jhep.2023.06.004. PMID:37390966.
10. Junge N, Hentschel H, Krebs-Schmitt D, Stalke A, Pfister ED, Hartleben B, et al. Mild Crigler-Najjar Syndrome with Progressive Liver Disease — A Multicenter Retrospective Cohort Study. Children (Basel). 2023;10(9):1431. doi:10.3390/children10091431. PMID:37761392.
11. Kovačić Perica M, Todorić I, Marčinković N, Džepina P, Aničić MN, Mrzljak A, Vuković J. Case report: Crigler-Najjar syndrome type 1 in Croatia — more than a one in a million: a case series. Front Pediatr. 2023;11:1276349. doi:10.3389/fped.2023.1276349. PMID:37928349.
12. Aronson SJ, Junge N, Trabelsi M, Kelmemi W, Hubert A, Brigatti KW, et al.; CureCN Consortium. Disease burden and management of Crigler-Najjar syndrome: Report of a world registry. Liver Int. 2022;42(7):1593–604. doi:10.1111/liv.15239. PMID:35274801.
13. National Institute for Health and Care Excellence. Jaundice in newborn babies under 28 days. Clinical guideline CG98. Published 19 May 2010; last updated 31 October 2023. [https://www.nice.org.uk/guidance/cg98](https://www.nice.org.uk/guidance/cg98). Accessed 7 October 2026.
14. Kemper AR, Newman TB, Slaughter JL, Maisels MJ, Watchko JF, Downs SM, et al. Clinical Practice Guideline Revision: Management of Hyperbilirubinemia in the Newborn Infant 35 or More Weeks of Gestation. Pediatrics. 2022;150(3):e2022058859. doi:10.1542/peds.2022-058859. PMID:35927462.
15. Sambati V, Laudisio S, Motta M, Esposito S. Therapeutic Options for Crigler-Najjar Syndrome: A Scoping Review. Int J Mol Sci. 2024;25(20):11006. doi:10.3390/ijms252011006. PMID:39456788.
16. Strauss KA, Ahlfors CE, Soltys K, Mazareigos GV, Young M, Bowser LE, et al. Crigler-Najjar Syndrome Type 1: Pathophysiology, Natural History, and Therapeutic Frontier. Hepatology. 2020;71(6):1923–39. doi:10.1002/hep.30959. PMID:31553814.
17. D'Antiga L, Beuers U, Ronzitti G, Brunetti-Pierri N, Baumann U, Di Giorgio A, et al. Gene Therapy in Patients with the Crigler-Najjar Syndrome. N Engl J Med. 2023;389(7):620–31. doi:10.1056/NEJMoa2214084. PMID:37585628.
18. U.S. Food and Drug Administration, Office of Orphan Products Development. Searchable Orphan Drug Designations and Approvals Database. Searches in the “Orphan Designation” field for “Crigler” (4 designations) and “hyperbilirubinemia” (1 designation), from 1 January 1983 to 7 October 2026, detailed format. [https://www.accessdata.fda.gov/scripts/opdlisting/oopd/](https://www.accessdata.fda.gov/scripts/opdlisting/oopd/). Accessed 7 October 2026.
19. European Medicines Agency. Orphan designations for Crigler-Najjar syndrome: EU/3/14/1338 (adeno-associated viral vector serotype 8 containing the human UGT1A1 gene, Genethon, 15 October 2014), EU/3/07/506, EU/3/16/1684, EU/3/14/1321 and EU/3/16/1772. [https://www.ema.europa.eu/en/medicines/human/orphan-designations/eu-3-14-1338](https://www.ema.europa.eu/en/medicines/human/orphan-designations/eu-3-14-1338). Accessed 7 October 2026.
20. European Medicines Agency. List of medicines currently in the PRIME scheme (EMA/521657/2016), version dated 23 September 2026; GNT0003 entry, eligibility granted 26 January 2023. [https://www.ema.europa.eu/en/documents/other/list-medicines-currently-prime-scheme_en.xlsx](https://www.ema.europa.eu/en/documents/other/list-medicines-currently-prime-scheme_en.xlsx). Accessed 7 October 2026.
21. Gammal RS, Court MH, Haidar CE, Iwuchukwu OF, Gaur AH, Alvarellos M, et al.; Clinical Pharmacogenetics Implementation Consortium. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for UGT1A1 and Atazanavir Prescribing. Clin Pharmacol Ther. 2016;99(4):363–9. doi:10.1002/cpt.269. PMID:26417955.
22. Hulshof EC, Deenen MJ, Nijenhuis M, Soree B, de Boer-Veger NJ, Buunk AM, et al. Dutch pharmacogenetics working group (DPWG) guideline for the gene–drug interaction between UGT1A1 and irinotecan. Eur J Hum Genet. 2023;31(9):982–7. doi:10.1038/s41431-022-01243-2. PMID:36443464.
23. Chauhan D, Dello Russo C, Thompson J, Hughes DA, Payne K, Tsakiroglou M, et al. UGT1A1 genotype testing for irinotecan: A guideline developed by the UK Centre of Excellence in Regulatory Science and Innovation in Pharmacogenomics (CERSI-PGx). Br J Clin Pharmacol. 2026;92(9):2986–3001. doi:10.1002/bcp.70659. PMID:42389808.
24. U.S. Food and Drug Administration. CAMPTOSAR (irinotecan hydrochloride injection) — prescribing information. NDA 020571. Version effective 12 May 2026. Accessed 7 October 2026.
25. MetabERN — European Reference Network for Hereditary Metabolic Disorders. Hôpital Antoine-Béclère, Service de Pédiatrie (AP-HP). [https://metab.ern-net.eu/hcp/hopital-antoine-beclere-service-de-pediatrie-ap-hp/](https://metab.ern-net.eu/hcp/hopital-antoine-beclere-service-de-pediatrie-ap-hp/). Accessed 7 October 2026.
26. Bernts LHP, Jones DEJ, Kaatee MM, Lohse AW, Schramm C, Sturm E, Drenth JPH. Position statement on access to care in rare liver diseases: advancements of the European reference network (ERN) RARE-LIVER. Orphanet J Rare Dis. 2019;14(1):169. doi:10.1186/s13023-019-1152-z. PMID:31287000.
27. European Commission, CORDIS. New research project CureCN aims to cure the ultra-rare Crigler-Najjar syndrome. 5 July 2018. [https://cordis.europa.eu/article/id/123204-new-research-project-curecn-aims-to-cure-the-ultrarare-criglernajjar-syndrome](https://cordis.europa.eu/article/id/123204-new-research-project-curecn-aims-to-cure-the-ultrarare-criglernajjar-syndrome). Accessed 7 October 2026.
28. Crigler JF Jr, Najjar VA. Congenital familial nonhemolytic jaundice with kernicterus. Pediatrics. 1952;10(2):169–80. PMID:12983120.
29. Arias IM, Gartner LM, Cohen M, Ben Ezzer J, Levi AJ. Chronic nonhemolytic unconjugated hyperbilirubinemia with glucuronyl transferase deficiency. Clinical, biochemical, pharmacologic and genetic evidence for heterogeneity. Am J Med. 1969;47(3):395–409. doi:10.1016/0002-9343(69)90224-1. PMID:4897277.
30. Kadakol A, Ghosh SS, Sappal BS, Sharma G, Chowdhury JR, Chowdhury NR. Genetic lesions of bilirubin uridine-diphosphoglucuronate glucuronosyltransferase (UGT1A1) causing Crigler-Najjar and Gilbert syndromes: correlation of genotype to phenotype. Hum Mutat. 2000;16(4):297–306. doi:10.1002/1098-1004(200010)16:4<297::AID-HUMU2>3.0.CO;2-Z. PMID:11013440.
31. Jesus Sá C, Vilar Queirós R, Pessoa T, Santos S, Castilho S. Between Crigler-Najjar Syndrome Type II and Gilbert Syndrome: Expanding the Spectrum of Uridine Diphosphate Glucuronosyltransferase 1A1 (UGT1A1)-Related Hyperbilirubinemia. Cureus. 2026;18(1):e100926. doi:10.7759/cureus.100926. PMID:41658774.
32. ClinVar. Records for alleles UGT1A1*28 (VCV000012275), *36 (VCV000549830), *37 (VCV000549829), *6 (VCV000012280) and variants c.686C>A and c.722_723del. National Center for Biotechnology Information. [https://www.ncbi.nlm.nih.gov/clinvar/](https://www.ncbi.nlm.nih.gov/clinvar/). Accessed 7 October 2026.
33. Petit FM, Bézieau S, Gajdos V, Parisot F, Scoul C, Capel L, et al. The Tunisian population history through the Crigler-Najjar type I syndrome. Eur J Hum Genet. 2008;16(7):848–53. doi:10.1038/sj.ejhg.5201989. PMID:18197191.
34. Orphanet. Dubin-Johnson syndrome. ORPHA:234. Reviewers: Do Cao J, Labrune P (MetabERN). Last updated February 2026. [https://www.orpha.net/en/disease/detail/234](https://www.orpha.net/en/disease/detail/234). Accessed 7 October 2026.
35. Dhawan A, Lawlor MW, Mazariegos GV, McKiernan P, Squires JE, Strauss KA, et al. Disease burden of Crigler-Najjar syndrome: Systematic review and future perspectives. J Gastroenterol Hepatol. 2020;35(4):530–43. doi:10.1111/jgh.14853. PMID:31495946.
36. Saudubray JM, Garcia-Cazorla À. Clinical Approach to Inborn Errors of Metabolism in Paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, sections 1.5 and 1.5.1.1. doi:10.1007/978-3-662-63123-2_1.
37. Firth HV, Hurst JA. Deafness in early childhood. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 122.
38. Jaber M, Abuzahra A, Karajat AK, Abu Shama'a B, Hunyhen LK, Baradia L, et al. Case Report: Therapeutic plasma exchange for severe unconjugated hyperbilirubinemia with bilirubin-induced neurologic dysfunction in an adolescent with Crigler-Najjar syndrome type II. Front Pediatr. 2026;14:1845433. doi:10.3389/fped.2026.1845433. PMID:42676547.
39. Kaplan M, Renbaum P, Levy-Lahad E, Hammerman C, Lahad A, Beutler E. Gilbert syndrome and glucose-6-phosphate dehydrogenase deficiency: a dose-dependent genetic interaction crucial to neonatal hyperbilirubinemia. Proc Natl Acad Sci U S A. 1997;94(22):12128–32. doi:10.1073/pnas.94.22.12128. PMID:9342374.
40. Wang J, Yin J, Xue M, Lyu J, Wan Y. Roles of UGT1A1 Gly71Arg and TATA promoter polymorphisms in neonatal hyperbilirubinemia: A meta-analysis. Gene. 2020;736:144409. doi:10.1016/j.gene.2020.144409. PMID:32007587.
41. Li H, Zhang P. UGT1A1*28 gene polymorphism was not associated with the risk of neonatal hyperbilirubinemia: a meta-analysis. J Matern Fetal Neonatal Med. 2021;34(24):4064–71. doi:10.1080/14767058.2019.1702962. PMID:31818155.
42. Passuello V, Puhl AG, Wirth S, Steiner E, Skala C, Koelbl H, Kohlschmidt N. Pregnancy outcome in maternal Crigler-Najjar syndrome type II: a case report and systematic review of the literature. Fetal Diagn Ther. 2009;26(3):121–6. doi:10.1159/000238122. PMID:19752526.
43. Gajdos V, Petit F, Trioche P, Mollet-Boudjemline A, Chauveaud A, Myara A, et al. Successful pregnancy in a Crigler-Najjar type I patient treated by phototherapy and semimonthly albumin infusions. Gastroenterology. 2006;131(3):921–4. doi:10.1053/j.gastro.2006.07.005. PMID:16952560.
44. Fawaz R, Baumann U, Ekong U, Fischler B, Hadzic N, Mack CL, et al. Guideline for the Evaluation of Cholestatic Jaundice in Infants: Joint Recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr. 2017;64(1):154–68. doi:10.1097/MPG.0000000000001334. PMID:27429428.
45. Amendola M, Squires JE. Pediatric Genetic Cholestatic Liver Disease Overview. In: GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published 15 September 2022; revised 25 May 2023. NBK584020.
46. Wevers RA, Mayatepek E, Walker V. Disorders of Peptide and Amine Metabolism, section 31.3 Disorders of Glutathione Metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 31, p. 571–85, section 31.3.6. doi:10.1007/978-3-662-63123-2_31.
47. Walter JH. Disorders of Galactose Metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 14, section 14.1.1. doi:10.1007/978-3-662-63123-2_14.
48. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Fructose-1,6-diphosphatase deficiency. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 49, p. 372.
49. Ribeiro EM, coord. Síndrome do filho de mãe diabética. In: Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 184.
50. Sardh E, Harper P. Disorders of Haem Biosynthesis. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 33, section 33.2.3. doi:10.1007/978-3-662-63123-2_33.
51. Al-Hussaini A, AlSaleem B, AlHomaidani H, Asery A, Alruwaithi M, Alameer M, et al. Clinical, Biochemical, and Molecular Characterization of Neonatal-Onset Dubin-Johnson Syndrome in a Large Case Series From the Arabs. Front Pediatr. 2021;9:741835. doi:10.3389/fped.2021.741835. PMID:34858902.
52. Al-Hussaini A, Asery A, Alharbi O. Urinary coproporphyrins as a diagnostic biomarker of Dubin-Johnson syndrome in neonates: A diagnostic pathway is proposed. Saudi J Gastroenterol. 2023;29(3):183–90. doi:10.4103/sjg.sjg_480_22. PMID:37313948.
53. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Appendix: Differential diagnosis of clinical phenotypes — Cholestatic jaundice. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 818.
54. Strauss KA, Robinson DL, Vreman HJ, Puffenberger EG, Hart G, Morton DH. Management of hyperbilirubinemia and prevention of kernicterus in 20 patients with Crigler-Najjar disease. Eur J Pediatr. 2006;165(5):306–19. doi:10.1007/s00431-005-0055-2. PMID:16435131.
55. Grifols Therapeutics LLC. Plasbumin-25 (Albumin [Human]) 25%: prescribing information. DailyMed setid 2eeb6b69-2013-c0a2-5e2f-0365114d41b4; version published 6 Dec 2022 [accessed 26 Oct 2026]. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=2eeb6b69-2013-c0a2-5e2f-0365114d41b4
56. Dr. Falk Pharma Portugal. Ursofalk 250 mg cápsulas (ácido ursodesoxicólico): resumo das características do medicamento. Lisbon: INFARMED, I.P.; text approved 27 Jun 2024 [accessed 26 Oct 2026]. Available from: https://extranet.infarmed.pt/INFOMED-fo/
57. U.S. Food and Drug Administration. Ursodiol capsules (NDA019594): prescribing information. Silver Spring (MD): FDA; version of 19 Jan 2026 [accessed 26 Oct 2026].
58. Sandoz Farmacêutica. Calcioral 1250 mg comprimido para mastigar (carbonato de cálcio): resumo das características do medicamento. Lisbon: INFARMED, I.P.; text approved 2 Aug 2024 [accessed 26 Oct 2026]. Available from: https://extranet.infarmed.pt/INFOMED-fo/
59. Hafkamp AM, Nelisse-Haak R, Sinaasappel M, Oude Elferink RP, Verkade HJ. Orlistat treatment of unconjugated hyperbilirubinemia in Crigler-Najjar disease: a randomized controlled trial. Pediatr Res. 2007;62(6):725–30. doi:10.1203/PDR.0b013e3181598cc5. PMID:17957158.
60. BIAL – Portela & Cª. Bialminal 100 mg comprimidos (fenobarbital): resumo das características do medicamento. Lisbon: INFARMED, I.P.; text approved 9 Jun 2022 [accessed 26 Oct 2026]. Available from: https://extranet.infarmed.pt/INFOMED-fo/
61. Bluepharma Genéricos. Luminaletas 15 mg comprimidos (fenobarbital): resumo das características do medicamento. Lisbon: INFARMED, I.P. [accessed 26 Oct 2026]. Available from: https://extranet.infarmed.pt/INFOMED-fo/
62. U.S. Food and Drug Administration. Sezaby (phenobarbital sodium) for injection (NDA215910): prescribing information. Silver Spring (MD): FDA; version of 23 Dec 2025 [accessed 26 Oct 2026].
63. Teimoury S, Beyzaei Z, Shamsaeefar A, Kazemi K, Nikeghbalian S, Malekhosseini SA, Geramizadeh B. Liver Transplantation Outcomes in Crigler-Najjar Syndrome in Iran: A Single-Center Retrospective Cohort Study Over 20 Years. Health Sci Rep. 2026;9(5):e72393. doi:10.1002/hsr2.72393. PMID:42057867.
64. Fox IJ, Chowdhury JR, Kaufman SS, Goertzen TC, Chowdhury NR, Warkentin PI, et al. Treatment of the Crigler-Najjar syndrome type I with hepatocyte transplantation. N Engl J Med. 1998;338(20):1422–6. doi:10.1056/NEJM199805143382004. PMID:9580649.
65. ClinicalTrials.gov. NCT03466463. A Phase I/II, Open Label, Study to Evaluate Safety and Efficacy of an Intravenous Injection of GNT0003 (AAV Vector Expressing the UGT1A1 Transgene) in Patients With Severe Crigler-Najjar Syndrome Requiring Phototherapy. Sponsor: Genethon. [https://clinicaltrials.gov/study/NCT03466463](https://clinicaltrials.gov/study/NCT03466463). Accessed 7 October 2026.
66. ClinicalTrials.gov. NCT06518005. An Open-label, Phase 2 Trial to Evaluate the Efficacy and Safety of a Single Intravenous Administration of GNT0003 Following Imlifidase Pre-treatment in Adult Participants With Severe Crigler-Najjar Syndrome Requiring Daily Phototherapy and Presenting Pre-existing Anti-AAV8 Antibodies. Sponsor: Genethon. [https://clinicaltrials.gov/study/NCT06518005](https://clinicaltrials.gov/study/NCT06518005). Accessed 7 October 2026.
67. ClinicalTrials.gov. NCT06641154. A Phase I/II, Open Label, Study to Evaluate Safety and Efficacy of an Intravenous Injection of GT-UGT1A1-AAV8-02 in Patients with Crigler-Najjar Syndrome Type I Requiring Phototherapy (AlphaCN). [https://clinicaltrials.gov/study/NCT06641154](https://clinicaltrials.gov/study/NCT06641154). Accessed 7 October 2026.
68. ClinicalTrials.gov. NCT03223194. VALENS: A Phase 1/2 Clinical Study to Evaluate the Safety and Preliminary Efficacy of AT342, an AAV8-Delivered Gene Transfer Therapy in Crigler-Najjar Syndrome Subjects Aged 1 Year and Older. Sponsor: Audentes Therapeutics. Status: terminated by sponsor decision. [https://clinicaltrials.gov/study/NCT03223194](https://clinicaltrials.gov/study/NCT03223194). Accessed 7 October 2026.
69. ClinicalTrials.gov. NCT01765283. A Prospective, Open Label, Multicenter, Partially Randomized, Safety Study of One Cycle of Promethera HepaStem in Urea Cycle Disorders and Crigler-Najjar Syndrome Paediatric Patients. [https://clinicaltrials.gov/study/NCT01765283](https://clinicaltrials.gov/study/NCT01765283). Accessed 7 October 2026.
70. ClinicalTrials.gov. NCT00461799. Orlistat Treatment of Crigler-Najjar Disease. Sponsor: University Medical Center Groningen. [https://clinicaltrials.gov/study/NCT00461799](https://clinicaltrials.gov/study/NCT00461799). Accessed 7 October 2026.
71. Genethon, Hansa Biopharma. Data from GNT-018-IDES trial supports feasibility of imlifidase as pretreatment in gene therapy treatment for patients with Crigler-Najjar syndrome who are immune to AAV. Press release, 10 October 2025. The case was later published as an opinion article: Do Cao J, Gao-Desliens J, Valent A, Knuchel-Legendre N, Colpin C, Guyon H, et al. Imlifidase enables AAV gene therapy in a seropositive Crigler-Najjar patient: on the verge of a paradigm shift. Mol Ther. 2026;34(5):2571–2. doi:10.1016/j.ymthe.2026.04.018. PMID:42049029. [https://www.biospace.com/press-releases/data-from-gnt-018-ides-trial-supports-feasibility-of-imlifidase-as-pretreatment-in-gene-therapy-treatment-for-patients-with-crigler-najjar-syndrome-who-are-immune-to-aav](https://www.biospace.com/press-releases/data-from-gnt-018-ides-trial-supports-feasibility-of-imlifidase-as-pretreatment-in-gene-therapy-treatment-for-patients-with-crigler-najjar-syndrome-who-are-immune-to-aav). Accessed 7 October 2026.
72. U.S. Food and Drug Administration. ONIVYDE (irinotecan liposome injection) — prescribing information. NDA 207793. Version effective 3 March 2025. Accessed 7 October 2026.
73. European Medicines Agency. ONIVYDE pegylated liposomal — summary of product characteristics (EPAR product information). [https://www.ema.europa.eu/en/medicines/human/EPAR/onivyde-pegylated-liposomal](https://www.ema.europa.eu/en/medicines/human/EPAR/onivyde-pegylated-liposomal). Accessed 7 October 2026.
74. U.S. Food and Drug Administration. TRODELVY (sacituzumab govitecan-hziy) — prescribing information. BLA 761115. Version effective 24 June 2026. Accessed 7 October 2026.
75. European Medicines Agency. Trodelvy — summary of product characteristics (EPAR product information). [https://www.ema.europa.eu/en/medicines/human/EPAR/trodelvy](https://www.ema.europa.eu/en/medicines/human/EPAR/trodelvy). Accessed 7 October 2026.
76. Dutch Pharmacogenetics Working Group (KNMP). Annotation of DPWG Clinical Guideline for sacituzumab govitecan and UGT1A1. ClinPGx PA166436801, content dated 27 January 2025, published 27 April 2026. [https://www.clinpgx.org/guidelineAnnotation/PA166436801](https://www.clinpgx.org/guidelineAnnotation/PA166436801). Accessed 7 October 2026.
77. U.S. Food and Drug Administration. BELEODAQ (belinostat) for injection — prescribing information. NDA 206256. Version effective 7 April 2025. Accessed 7 October 2026.
78. U.S. Food and Drug Administration. REYATAZ (atazanavir) — prescribing information. NDA 021567 and NDA 206352. Version effective 5 December 2024. Accessed 7 October 2026.
79. European Medicines Agency. Reyataz — summary of product characteristics (EPAR product information). [https://www.ema.europa.eu/en/medicines/human/EPAR/reyataz](https://www.ema.europa.eu/en/medicines/human/EPAR/reyataz). Accessed 7 October 2026.
80. U.S. Food and Drug Administration. TASIGNA (nilotinib) capsules — prescribing information. NDA 022068. Version effective 16 December 2025. Accessed 7 October 2026.
81. U.S. Food and Drug Administration. VOTRIENT (pazopanib) tablets — prescribing information. NDA 022465. Version effective 16 December 2025. Accessed 7 October 2026.
82. Swen JJ. UGT1A1 and irinotecan: Ready for routine pharmacogenetic care. Br J Clin Pharmacol. 2026;92(9):2983–5. doi:10.1002/bcp.70692. PMID:42437664.
83. Dutch Pharmacogenetics Working Group (KNMP). Annotation of DPWG Clinical Guideline for atazanavir and UGT1A1. ClinPGx PA166411701, created 27 June 2025. [https://www.clinpgx.org/guidelineAnnotation/PA166411701](https://www.clinpgx.org/guidelineAnnotation/PA166411701). Accessed 7 October 2026.
84. Heersche N, Peeters SLJ, Böhm D, Böhringer S, Guiljam R, Hulshof EC, et al. Real-World Impact of UGT1A1 Genotype-Guided Irinotecan Dosing on Severe Toxicity and Hospitalization: A Multicenter Study. J Natl Compr Canc Netw. 2026;24(7):e267011. doi:10.6004/jnccn.2026.7011. PMID:42276133.
85. National Organization for Rare Disorders. Crigler-Najjar Association / King's Way Foundation. [https://rarediseases.org/organizations/crigler-najjar-association-kings-way-foundation/](https://rarediseases.org/organizations/crigler-najjar-association-kings-way-foundation/). Accessed 7 October 2026.