Clinical observations disorders [5]. Inborn errors of metabolism cause only 1% of NIHF. The risk of recurrence in subsequent pregnancies is high [6]. Only a few cases of GM1 gangliosidosis presenting with NIHF have been reported so far; their features are summarized in Table I. It is interesting to note that our patient is the only one to have a very early onset of symptoms (23 wk of gestation) and to have a thorax involvement. GM1 gangliosidosis does not usually have a prenatal presentation; when it happens, the most common finding is fetal or neonatal ascites, although the reason is still unclear. Storage material in Kupffer cells, sinusoidal obstruction and subsequent portal hypertension, or hypoproteinaemia due to hepatocellular dysfunction, have been hypothesized as possible explanations. However, in our case, liver ultrasound and hepatic function were normal. Abu Dalu considered the possibility that lymphatic vessels may enhance permeability resulting in fluid extravasation into the peritoneal cavity. In addition, our baby also had pleural effusion that could not be justified by these hypotheses. An interesting aspect of our patient is the presence of hyponatraemia and hypernatriuria, suggesting a tubular dysfunction, not previously described in GM1 gangliosidosis. The real incidence of metabolic diseases as a cause of fetalis hydrops is unknown. In the past they were often not suspected and many cases were not detected. We suggest that the possible association between inborn errors of metabolism and antenatal ascites should be considered, in order to offer genetic 1849 counselling due to the high recurrence risk and the availability of early antenatal diagnosis. References [1] O’Brien JS, Stern MB, Landing BH, O’Brien JK, Donnell GN. Generalized gangliosidosis. Am J Dis Child 1965;109:338. [2] O’Brien JS. The gangliosidosis. In: Stambury JB, Wyngaarden JB, Fredrickson DS, Goldstein JL, Brown MS, editors. The metabolic basis of inherited disease. 5th ed. New York: McGraw Hill; 1983. p 945–69. [3] Singer HS, Schafer IA. White cell beta-galactosidase activity. N Engl J Med 1970;282:571. [4] Hutchison AA, Drew JH, Yu VYH, Williams ML, Fortune DW, Beischer NA. Nonimmunologic hydrops fetalis: a review of 61 cases. Obstet Gynecol 1982;59:347–52. [5] Heffner LJ, Roberts DJ, Nanowitz M. A premature newborn infant with congenital ascites. New Engl J Med 1997;337:260–7. [6] Norton ME. Nonimmune hydrops fetalis. Semin Perinatol 1994;18:321–32. [7] Abu-Dalu KI, Tamary H, Livni N, Rivkind AI, Yatziv S. GM1 gangliosidosis presenting as neonatal ascites. J Pediatr 1982; 100:940–3. [8] Guillan JE, Lowden JA, Gaskin K, Cutz E. Congenital ascites as a presenting sign of lysosomal storage disease. J Pediatr 1984;104:225–31. [9] Vajro P, Strisciuglio P, Fontanella A, De Vincenzo A. Neonatal ascites disclosing GM1 gangliosidosis. Arch Fr Pediatr 1990; 47:544. [10] Bonduelle M, Lissens W, Goossens A, De Catte L, Foulon W, Denis R, Jauniaux E, Liebaers I. Lysosomal storage diseases presenting as transient or persistent hydrops fetalis. Genet Couns 1991;2:227. [11] Tasso MJ, Martinez-Gutierrez A, Carrascosa C, Vazquez S, Tebar R. GM1-gangliosidosis presenting as nonimmune hydrops fetalis: a case report. J Perinat Med 1996;24:445–9. Cirrhosis in an infant heterozygous for classical citrullinaemia FATIH SÜHEYL EZGÜ1, LEYLA TÜMER1, BUKET DALGIÇ2, ALEV HASANOĞLU1, KEIKO KOBAYASHI3 & TAKEYORI SAHEKI3 1 Department of Paediatric Metabolism and Nutrition, Gazi University Faculty of Medicine, Ankara, Turkey, 2Department of Paediatric Gastroenterology, Gazi University Faculty of Medicine, Ankara, Turkey, and 3Department of Molecular Metabolism and Biochemical Genetics, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan Abstract Classical citrullinaemia is caused by the inherited deficiency of argininosuccinate synthetase. Although varying degrees of liver involvement have been observed in urea cycle defects, including classical citrullinaemia, the co-existence of liver failure in a patient heterozygous for the disease has not been reported before. A female infant was investigated to find out the aetiology of early infantile liver failure. She was later found to be a heterozygote for the G390R mutation found in severe citrullinaemia patients. Conclusion: Classical citrullinaemia is a phenotypically heterogeneous disease, and observations for signs of its presence should be made even in heterozygotes. Key Words: Infantile, urea cycle disorder, cirrhosis Correspondence: F. S. Ezgü, Gazi University Hospital, Department of Paediatrics, 10th Floor, Besevler, 06500, Ankara, Turkey. Tel: +90 312 2026027. Fax: +90 312 215 01 43. E-mail: fatihezgu@isbank.net.tr (Received 11 December 2004; revised 11 December 2004; accepted 25 February 2005) ISSN 0803-5253 print/ISSN 1651-2227 online # 2005 Taylor & Francis DOI: 10.1080/0803525051003078 1850 Clinical observations Introduction Hepatic failure in infancy is a rare but serious event that results in the death of most children affected. Among the most common causes are congenital biliary abnormalities, infections, drugs, toxins and inborn errors [1,2]. Inherited deficiency of the argininosuccinate synthetase (ASS; EC 6.3.4.5) gene causes a recessively transmitted disease: classical citrullinaemia (CTLN1; MIM 215700). The neonatal form of the disease generally involves hyperammonaemic coma in the first days of life and leads to neurological sequelae in survivors [3–5]. However, like many other genetic diseases, citrullinaemia is likely to be phenotypically heterogeneous [5]. Although varying degrees of liver involvement have been described [6] and cirrhosis due to classical citrullinaemia has recently been reported [7], the occurrence of early infantile liver failure in an infant heterozygous for the G390R mutation found in the severe citrullinaemia patients [8] has not been reported before. Case report A 1-d-old female infant was referred to the Department of Paediatric Metabolism, Gazi University Hospital, in order to rule out CTLN1, because her newborn brother had been suspected of having the disease just before his death 7 y earlier. Her brother, after an uneventful delivery, had suffered from a hyperammonaemic coma 10 d after birth and had increased urine orotic acid, plasma citrulline and glutamine levels (Table I). The parents were cousins, and they refused prenatal diagnosis for the present case. The patient was born after an uneventful pregnancy. The mother did not use any medications during pregnancy or the breastfeeding period. The patient’s initial physical examination was completely in the normal range. On the second day after birth,her blood citrulline level was found to be 22 mmol/l (normal 8–35) with tandem mass spectrometry. Neither orotic aciduria nor hyperammonaemia were detected. Liver function tests including alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma glutamyl transpeptidase (GGT), alkaline phosphatase (ALP) and bilirubin levels were also normal. The same investigations were repeated on days 7 and 15, and the citrulline level was found to be 35 and 32.5 mmol /l, respectively. The patient was admitted to our department on the 35th day of life with jaundice. Physical examination revealed a hepatomegaly of 2 cm below the costal margin besides generalized jaundice. Liver function tests revealed: AST 95 IU/l (normal 15–60); ALT 82 IU/l (normal 13–45); ALP 520 IU/l (normal 150–420); GGT 232 IU/l (normal 4–120); total bilirubin 182.5 mmol/l (normal 5171); direct bilirubin 61.9 mmol/l (normal 53.4); prothrombin time 14 s (normal 10.6–11.4); activated partial thromboplastin time 36 s (normal 24–36). The results of blood amino acid, ammoniac and urinary organic acid investigations were in the normal range, with citrulline level being in the upper limit of the normal range (34.5 mmol/l). All tests regarding the aetiology of the liver impairment, including sweat chloride and a1 antitrypsin, lysosomal enzymes, very-long-chain fatty acids, reducing substance and organic acids in the urine, thyroid function tests, markers of viral hepatitis, bone marrow aspiration, echocardiography, Doppler ultrasonography, hepatocyte uptake and excretion on scintigraphy in the 4th month of life, were normal. Pathological examination of the liver biopsy performed in the 8th month revealed hepatocellular damage and mild cholestasis. The patient was hospitalized 10 mo after birth because of gastrointestinal bleeding. In the physical examination, the liver was palpated 8 cm and the spleen was palpated 5 cm below the right and left costal margin, respectively. The prominence of the superficial veins on the abdominal wall and generalized jaundice were also noticed. Ascite was noticed by ultrasonography. Laboratory examinations revealed: AST 461 IU/l (normal 15–60); ALT 211 IU/l (normal 13–45); total bilirubin 345.42 mmol/l (normal 5171); direct bilirubin 246.24 mmol/l (normal 53.4); prothrombin time 22.4 s (normal 10.6–11.4); activated partial thromboplastin time 42.7 s (normal 24–36); blood ammonia level 185.84 mmol/l (normal 5125). Oesophageal varices were noticed in barium graphy. The patient was treated with erythrocyte and fresh frozen plasma infusions, intramuscular vitamin K, oral spironolactone, propranolol, ursodeoxycholic acid and a high-energy diet rich in medium-chain triglycerides to which she responded. A second biopsy was performed when the patient was 1 y old, which revealed cirrhosis. In order to rule out citrin deficiency, SLC25A13 gene mutations were studied in addition to mutations for the ASS gene. Although the mutational analysis of the SLC25A13 gene was negative, RT-PCR/direct sequencing of the RNA extracted from the fibroblasts revealed heterozygosity for G390R mutation in the ASS gene [8] in the patient as well as the parents. The citrulline and arginine levels of the parents and the patient at 24 mo of age are shown in Table I. Discussion The extensive work-up to clear the aetiology of liver failure in our patient failed, except for the finding of heterozygosity for CTLN1. A literature review of the clinical and biochemical parameters of the cases with CTLN1 indicates that significant phenotypical variation exists, especially for the cases of Turkish and Mediterranean origin [3–5]. In these cases, the clinical picture ranges from asymptomatic phenotypes with mild elevation in blood citrulline and no/mild hyperammonaemia to death [3–5]. In some genetic diseases with such a wide spectrum of clinic heterogenicity, Clinical observations Table I. Citrulline, arginine, glutamine (mmol/l) and orotic acid (mmol/mol creatinine) levels of the patient, her brother and the parents. Normal Citrulline 8–34 Arginine 7–128 Glutamine 402–776 Orotic acid 0.02–3.6 CTLN1 Patient Patient brother (35 d) (24 mo) Mother Father 1772 103 2006 21 34.5 38.7 486 u.d. 35.5 44.1 552 u.d. 40.6 81.6 n.m. u.d. 35.3 17.2 n.m. u.d. n.m.: not measured; u.d.: undetectable. some clinical or biochemical signs could be present in heterozygotes. However, to date, the only reported sign in heterozygotes for ASS mutation is mildly elevated fasting plasma citrulline levels [9]. Another genetically different disease that could lead to infantile liver failure with intrahepatic cholestasis is citrin deficiency [10]. The genetic origin was firstly described by Kobayashi et al. [11] and is caused by mutations of the SLC25A13 gene that lead to a defect of citrin, a mitochondrial aspartate glutamate carrier protein [12]. It is also characterized by the transient elevation of citrulline levels, hypoproteinaemia, galactosaemia and hypoglycaemia in the neonatal period, as well as a clinical picture of cholestatic liver failure [10,12]. However, in contrast to the clinical course of our patient, most patients with citrin deficiency show ameliorated symptoms by about 12 mo without special treatment [12]. This disorder was excluded in our case as no mutations were found in the SLC25A13 gene. Liver biopsy findings of patients with urea cycle defects range from normal to a broad spectrum of changes [6]. Although the light microscopic findings of the liver biopsy specimen of our patient did not point to a specific pathology, there were many similarities with previously described findings in patients with CTLN1, which included an increase in fibrous tissue, cholestasis and focal hepatocellular necrosis [6,7]. Generally the contribution of hyperammonaemia or hypercitrullinaemia to liver damage is suspected in these patients. We never observed hyperammonaemia in our patient, but the citrulline levels were slightly over the upper limit. The mutation G390R is one of the frequent mutations causing CTLN1 in the homozygote form, especially in Turkey. Compound heterozygotes may show milder phenotypes [5]. To our knowledge, this report is the first to describe the co-existence of heterozygosity for CTLN1 and infantile liver failure. Although the previous sibling, who seemed to be homozygous for 1851 the G390R mutation, presented with a full clinical picture including liver failure and, to our knowledge, there exists only one female patient in the literature with the G390R/G117D genotype who presented with Reye-like syndrome [4] and a Turkish infant homozygous for a G4C substitution at position 1 of IVS 15 of the ASS gene who presented with early cirrhosis [7], it is difficult to draw a conclusion about the pathophysiology or aetiology of the liver failure in our heterozygote case. Furthermore, it is very difficult to guess whether the hepatic failure of our case is somehow related to the heterozygosity for the G390R mutation which, in the homozygote form, causes a severe clinical expression [4,5,8]. Further observations are obviously needed. References [1] Treem WR. Fulminant hepatic failure in children. J Pediatr Gastroenterol Nutr 2002;35:S33–8. [2] Burdelski M, Nolkemper D, Ganschow R, Sturm E, Malago M, Rogiers X, et al. Liver transplantation in children: long-term outcome and quality of life. Eur J Pediatr 1999;158 Suppl 2:S34–42. [3] Tokatlı A, Coşkun T, Özalp İ. Citrullinemia. Clinical experience with 23 cases. Turk J Pediatr 1998;40:185–93. [4] Vilaseca MA, Kobayashi K, Briones P, Lambruschini N, Campistol J, Tabata A, et al. Phenotype and genotype heterogeneity in Mediterranean citrullinemia. Mol Genet Metab 2001;74:396–8. [5] Gao HZ, Kobayashi K, Tabata A, Tsuge H, Iijima M, Yasuda T, et al. Identification of 16 novel mutations in the argininosuccinate synthetase gene and genotype-phenotype correlation in 38 classical citrullinemia patients. Hum Mutat 2003;22:24–34. [6] Zamora SA, Pinto A, Scott RB, Parsons HG. Mitochondrial abnormalities of liver in two children with citrullinaemia. J Inherit Metab Dis 1997;20:509–16. [7] Güçer Ş, AŞan E, Atilla P, Tokatlı A, Çağlar M. Early cirrhosis in a patient with type I citrullinemia. J Inherit Metab Dis 2004;27:541–2. [8] Kobayashi K, Jackson MJ, Tick DB, O’Brien WE, Beaudet AL. Heterogeneity of mutations in argininosuccinate synthetase causing human citrullinemia. J Biol Chem 1990;265:11361–7. [9] Clemens PC, Plettner C. A non-enzymatic method for identification of citrullinemia heterozygotes. Clin Genet 1989; 35:468–9. [10] Tamamori A, Okano Y, Ozaki H, Fujimoto A, Kajiwara M, Fukuda K, et al. Neonatal intrahepatic cholestasis caused by citrin deficiency: severe hepatic dysfunction in an infant requiring liver transplantation. Eur J Pediatr 2002;161:609–13. [11] Kobayashi K, Sinasac DS, Iijima M, Boright AP, Begum L, Lee JR, et al. The gene mutated in adult-onset type II citrullinaemia encodes a putative mitochondrial carrier protein. Nat Genet 1999;22:159–63. [12] Saheki T, Kobayashi K. Mitochondrial aspartate glutamate carrier (citrin) deficiency as the cause of adult-onset type II citrullinemia (CTLN2) and idiopathic neonatal hepatitis (NICCD). J Hum Genet 2002;47:333–41.