Neonatal cholestasis due to citrin deficiency: diagnostic pitfalls

1Department of Pediatrics, Nutrition and Metabolic Diseases, The Children’s Memorial Health Institute, Warsaw, Poland; 2Department of Medical Genetics, The Children’s Memorial Health Institute, Warsaw, Poland; 3Department of Medical Genetics, Medical University of Warsaw, Warsaw, Poland; 4Department of Pediatrics, School of Medicine in Katowice, Medical University of Silesia, Katowice, Poland; 5Department of Biochemistry, Radioimmunology and Experimental Medicine, The Children’s Memorial Health Institute, Warsaw, Poland; 6Department of Gastroenterology, Hepatology, Feeding Disorders and Pediatrics, The Children’s Memorial Health Institute, Warsaw, Poland

We describe the clinical, biochemical, and molecular features of a Polish patient diagnosed with NICCD, and provide an overview of diagnostic pitfalls.

CASE REPORT
The patient was the third child of nonconsanguineous Polish parents born from an uneventful pregnancy at 38 weeks of gestation with a birth mass of 2 580 g (birth mass of two previous children was above 3 000 g). At the age of 5 weeks, he was referred to the local hospital due to prolonged jaundice and poor weight gain (body mass 3 020 g, <3rd centile). Cholestasis accompanied by slightly elevated serum transaminases, hypoalbuminemia and coagulopathy was diagnosed (Table 1). No history of hypo-/acholic stools was observed. Normal liver and spleen volume as well as the presence of gallbladder in abdominal ultrasound were observed. Infectious causes of cholestasis, including HBV, HCV, CMV, EBV, HIV, and Toxoplasma gondii infections, were excluded serologically. Alpha-1-antitrypsin (α1-AT) deficiency (based on normal serum α1-AT level), cystic fibrosis (based on normal serum immunoreactive trypsinogen level as a part of newborn screening programme), galactosemia (based on active galactose-1-phosphate uridylyltransferase) were excluded as well. Serum ammonia was not tested at this consultation. The improvement of cholestatic features was observed on ursodeoxycholic acid (UDCA) as well as fat-soluble vitamins treatment.
Due to raising parameters of cholestasis, noted at the age of 8 weeks, the child was transferred to our hospital. Patient's weight on admission was 3 800 g (still <3rd centile). The psychomotor development was normal. Results of laboratory analyses are shown in Table  1. Notably, several measurements revealed the serum ammonia to be mildly elevated. Serum alpha-fetoprotein (AFP) was noted elevated as well. Urinary gas chromatography-mass spectrometry (GC/MS) analysis revealed large quantities of 4-hydroxyphenylacetic acid and 4-hydroxyphenylpyruvate acid; succinylaceton was not detected, which allowed to exclude tyrosinemia. Protein N-hypoglycosylation in the form of increased a-, mono-, as well as di-sialotransferrin, was also detected. Tandem mass spectrometry analysis (MS/MS) of serum amino acids (AA) revealed high levels of citrulline (430 umol/l, reference range 1-55), threonine (475 umol/l, reference Due to presence of liver failure, the child was qualified to liver transplantation (LTx) procedure, and thus decision about molecular analysis (whole-exome sequencing, WES) was made. Results of all biochemical abnormalities were missed due to ongoing WES analysis.
During follow-up, at 6 months of age, the remission of cholestasis occured (Table 1). Serum AFP decreased while serum transaminases as well as ammonia level persisted as mildly elevated; oral lactulose was implemented. Normal serum ammonia was noted at the age of 9 months. In the whole-exome sequencing (WES) study, the patient was found to be homozygous for the novel splice site variant c.1453-2A>T in the SLC25A13 gene. The child was disqualified from LTx. Dietary treatment based on lipid and protein-rich low-carbohydrate was introduced.

DISCUSSION
Retrospectively, our patient showed almost all clinical and biochemical features of NICCD as summarized in Table 2. Unreported findings included protein N-hypoglycosylation and increased serum VLCFA. The main di- agnostic pitfall was that results of all biochemical abnormalities, especially high citrullinemia serum level, were missed due to ongoing WES analysis. The diagnostic approach of neonatal/infantile cholestasis is traditionally based on clinical, biochemical, imaging and histopathological findings (Sticova et al., 2018;Fawaz et al., 2017). Diagnosis of citrin deficiency is usually made on clinical and biochemical features including low birth weight, failure to thrive, early neonatal cholestasis with hypoalbuminemia, coagulopathy and elevated serum transaminases (Okano et al., 2019;He et al., 2019;Numakura et al., 2019;Chong et al., 2018;Tamamori et al., 2002;Ohura et al., 2007;Tazawa et al., 2004). Multiple aminoacidemia, including citrullinemia, are observed in majority of NICCD cases. In the presented case, amino acid and acylcarnitine profiles were normal in dried blood spots of newborn screening program. A relatively high blood citrulline level, observed at 3 months of patient's age, was characteristic for NIC-CD but it was missed due to ongoing WES analysis. The limited sensitivity of NICCD newborn screening may be related to an early blood sampling time (Tang et al., 2019).
A mild increase in the serum ammonia level was observed in our patient as well as in some reported NICCD patients. This is usually asymptomatic and the improvement on lactulose could be also misleading. Moreover, in the presented case, the serum ammonia level was analyzed for the first time at the age of 2 months. This is a cheap laboratory analysis and should be available at every neonatal unit. Monitoring of serum ammonia level reflects a valuable marker of liver metabolic disorders, therefore it should be carried out frequently.
However, there are some non-specific probably secondary biochemical features which were observed in our patient.
Increased serum VLCFA accumulation found in our patient was probably caused by a decreased peroxisomal activity in the liver tissue. It is not an unique finding; the ratio of C24:0/C22:0 as well as C26:0/C22:0 had been reported as increased in patients with liver insufficiency (Stradomska et al., 2013).
The abnormal serum isoelectric focusing (IEF) profile, indicative for protein N-hypoglycosylation, was also found. Since its introduction in 1984, serum transferrin IEF is still the method of choice for the diagnosis of N-glycosylation disorders with sialic acid deficiency (Jaeken et al., 1984). Some patients, including those with classic galactosemia (galactose-1-phosphate uridyltransferase deficiency), fructosemia (fructose 1-phosphate aldolase deficiency), severe liver impairment (personal observations), have secondarily an abnormal serum transferrin isoform profile. In our patient, this phenomenon was probably secondary to liver impairment in the course of citrin deficiency.
The final diagnosis of NICCD in our patient was established based on molecular analysis. The identificable causes of cholestasis have grown recently due to application of next-generation sequencing (NGS) technology (Sticova et al., 2018;Fawaz et al., 2017). Since the first reports of NICCD in Asian children, citrin deficency became now recognizable as a panethnic disorder.
NICCD is usually a self-limiting disease with remission of cholestasis seen between 6 and 12 months of age (Okano et al., 2019;He et al., 2019;Numakura et al., 2019;Chong et al., 2018;Tamamori et al., 2002;Ohura et al., 2007;Tazawa et al., 2004). The biochemical improvement seen on UDCA therapy could be also misleading for the definite diagnosis. However, there are few reports of patients who develop severe liver failure requiring liver transplantation (LTx) or died before LTx (Shigeta et al., 2010).

CONCLUSIONS
The paper presents a case of Polish NICCD patient presenting with low birth weight, failure to thrive, prolonged cholestatic jaundice with coagulopathy and hypoalbuminemia with normal results of MS/MS newborn screening but with high blood citrulline level observed at 3 months of age. Unreported findings included protein N-hypoglycosylation and increased serum very-longchain fatty acids (VLCFA), probably secondary to liver impairment. The main diagnostic pitfall was that results of all biochemical abnormalities, especially high serum citrulline level, were missed due to ongoing WES analysis.