Variations in the bilirubin UGT1A1 gene can induce different degrees of reduction in enzyme activity, resulting in unconjugated hyperbilirubinemia. The distinction between Crigler-Najjar syndrome type I or type II and GS is based also on serum bilirubin levels. These clinical entities are considered to be the spectrum of a single disorder with different grades of severity [3, 11, 12].
Our cases presented bilirubin values overlapping between GS and CNS type II, but the complete normalization of bilirubin makes GS more likely [2, 12]. In fact, in CNS type II no normalization of bilirubin is usually observed, owing to the severe lack of enzyme activity (usually < 10%) [5]. CNS type I (absence of the enzyme) was excluded on clinical and laboratory findings [12].
Even though in GS bilirubin amount usually ranges between 1 and 4 mg/dl [12], in some newborns higher values are described, above all when in association with other conditions such as G6PDH, pyloric stenosis or thalassemia [4, 9, 12]. In our patients no other associated factors were found, aside from breastfeeding.
Breastfeeding jaundice is a well-known entity, but its pathogenesis is still unclear [12, 13]. The initial hypothesis concerned the presence of substances capable of interfacing with hepatic conjugation in mother's milk (such as pregnanediol or an excess of non-esterified free fatty acids) [14]. Another hypothesis concerns the beta-glucuronidase enzyme, which is normally present in the brush border of the intestine and which is able to deconjugate bilirubin, increasing its reabsorption in the enterohepatic circulation [15]. Some studies have shown increased beta glucuronidase activity in breastfed infants compared to formula-fed ones [15]. Other studies focus on the presence of genetic variants of UGT1A1 that appear to be better inhibited by pregnanediol [16]. Further studies evaluated the presence of Gilbert syndrome mutation in breastfed jaundiced infants versus breastfed babies without jaundice and found no differences [12]. In conclusion, the combined effect of UGT 1A1 mutation together with the presence of breast milk may lead to a more severe jaundice in certain infants with predisposing factors [13]. Although prolonged unconjugated hyperbilirubinemia has been reported in formula-fed infants carrying UGT1A1 mutations, breastfed infants are more likely to present with this clinical picture [13]. Nevertheless, not all breastfed infants carrying UGT 1A1 mutations exhibit breast milk jaundice [7, 13]. In our cases breastfeeding jaundice was a possible explanation of the clinical picture, but a brief interruption of breastfeeding was not effective (Fig. 1).
The most severe phenotype of hyperbilirubinemia linked to UGT1A1 mutations is Crigler-Najjar syndrome type I, where the mutations result in inability to synthetize this enzyme. These patients will require lifelong phototherapy treatment until more definitive therapy, such as liver transplant, is proposed [11]. Because in Crigler-Najjar type II some UGT 1A1 activity is preserved, this disease has a more indolent course. The usual therapy for CNS II is phenobarbital(PB), which increases expression of the UGT1A1 gene in the liver and partially decreases the levels of bilirubin [11]. Gene therapy will probably be the final solution for CNS: animal models have been studied [17] and two trials on pediatric patients are ongoing (NCT03223194, NCT03466463).
PB response activity is delineated to a 290-bp distal enhancer module sequence (-3483/-3194) of the UGT1A1 gene called glucuronosyltransferase phenobarbital response enhancing motif (gtPBREM). Human constitutive active nuclear receptor (hCAR) is involved in activation of gtPBREM [18]. PB treatment results into translocation of cytoplasmic receptors like CAR into the nucleus; where it binds to retinoid X receptor and forms a heterodimer, which leads to activation of PB response enhancer element [18]. PB also stimulates hyperplasia of the endoplasmic reticulum (where UGT1A1 is located) [18]. Several studies on co-transfected HepG2 cells and mouse primary hepatocytes offered an excellent model for the examination of the responsiveness of the UGT1A1 to PB [11].
No specific treatment is usually necessary for GS, considering the benignity of the condition, although phenobarbital lowers serum bilirubin levels also in these patients [12]. In our cases phenobarbital apparently reduced bilirubin levels, but they maintained normal bilirubin levels also after the treatment. For this reason, even though the bilirubin levels showed a possible overlap with those described in CNS type II, we considered the cases as affected by an infantile and inducible phenotype of Gilbert’s Syndrome, homozygous for P364L mutation. The normalization of bilirubin probably was also obtained by the increase of UGT1A1 activity usually reached between the age of 6–12 weeks [7].
In other Chinese infants with unconjugated bilirubin level of ≥ 20 mg/dl the P364L variant was described (but in heterozygosity) [6, 10]. Hyperbilirubinemia is common in Chinese people and bilirubin levels are described as higher than in Caucasians, including newborns [1, 10]. Previous studies involving the Chinese population calculated the P364L carrier rate as 1.67% [2, 3] P364L polymorphism is presumably present in a certain percentage of Chinese newborns with severe jaundice.
In conclusion, homozygous P364L variant can be associated with severe neonatal unconjugated hyperbilirubinemia in Chinese infants, but jaundice can completely resolve, unlike what happens in CNS type II.
When evaluating a prolonged and severe neonatal jaundice, especially in Asiatic patients, clinicians must consider the possibility of the presence of P364L mutation in order to make an appropriate diagnosis and prognosis.