Neonatal Cholestatic Liver Disease in an Asian Patient With a Homozygous Mutation in the Oxysterol 7α‐hydroxylase Gene

We diagnosed a Taiwanese infant with oxysterol 7α-hydroxylase deficiency resulting from a homozygous mutation in the corresponding gene. We know of only 1 report of this inborn error of bile acid synthesis, by Setchell et al (1). The clinical and laboratory findings in our patient resembled those in patients with other inborn errors of bile acid synthesis. However, this rare form is difficult to treat with bile acid therapy because of its rapid progression to cirrhosis at an early stage. We present the clinical features, diagnostic evaluation, and treatment of our patient. CASE PRESENTATION AND METHODS Clinical History of Patient A male Taiwanese infant with a birth weight of 3050 g was delivered at term by cesarean section after an uneventful pregnancy. This was the mother's first pregnancy, and the parents were not consanguineous. At 5 months of age, the patient was referred to Taiwan University Hospital with a chief complaint of progressive jaundice over 2 weeks. Mild jaundice had been present since birth and had persisted for months, until his parents observed it worsening. Pruritus was absent. On physical examination, his growth and development were within the normal range. No dysmorphic features or cardiac murmurs were present. His neurological findings were normal. His stools were yellowish. Initial laboratory studies disclosed an AST concentration of 1080 U/L (normal <37); ALT, 358 U/L (< 41); alkaline phosphatase, 1116 U/L (60–220); and total/direct bilirubin, 6.2/4.3 mg/dL (<1.2/0.4). The serum γ-glutamyltransferase (GGT) concentration was 45 U/L (<52); total cholesterol, 143 mg/dL (130–220); and serum total bile acids were 4 μmol/L (<10). Abdominal sonography showed a visible gallbladder and hepatosplenomegaly. Liver biopsy specimen findings were consistent with cirrhosis, showing wide fibrotic bands at the portal areas. Bile duct proliferation was noted. Marked lobular disarray was prominent, accompanied by frequent giant cell transformation (Fig. 1). No coagulopathy was present and the serum albumin concentration was normal.FIG. 1: Liver histology at 5 mo of age (hematoxylin and eosin stain). Liver biopsy showed a marked increase in portal connective tissue with bile ductular proliferation and bridging fibrosis. Inset, A higher power view of the same biopsy showing multinucleated giant cell transformation.Although we treated the patient with ursodeoxycholic acid (UDCA, 15 mg · kg−1 · day−1), his total and direct bilirubin (10.0 and 7.6 mg/dL), AST (777 U/L), and ALT (389 U/L) values did not decline. His growth appeared stunted after 6 months of age. The patient was admitted at 9 months to a nearby medical center in southern Taiwan to be prepared for a liver transplantation, but his liver function deteriorated after an episode of infection. The patient died of chronic cholestatic liver failure when he was 11 months old. We could not administer cholic and/or chenodeoxycholic acids because these drugs are not available in Taiwan. Qualitative and Quantitative Bile Acid Analysis Serum and urine samples were collected and stored at −25°C until analysis. The concentrations of individual bile acids in the urine were corrected for his creatinine (Cr) concentration and expressed as micromoles per millimoles of Cr. After we synthesized some specific unusual bile acids—such as 3β-hydroxy-Δ5(2), 3-oxo-Δ4(3), and allo-bile acids (4)—that are found in inborn errors of bile acid synthesis, an analysis of the bile acids in his urine and serum was undertaken by gas chromatography-mass spectrometry (GC-MS) using selected ion monitoring of the characteristic fragments of the methylester-dimethylethylsilyl ether-methoxime derivatives of the bile acids as described previously (5). Before GC-MS, the samples were prepared by enzymatic hydrolysis (cholylglycine hydrolase, 30 U) and solvolysis (sulfatase, 150 U; Sigma Chemical, St Louis, MO). Genetic Analysis With informed consent, blood was collected from the patient and his parents, as well as 102 healthy individuals. Genomic DNA was extracted from peripheral leukocytes using a QIAamp Mini Kit (Qiagen, Hilden, Germany). Polymerase chain reaction (PCR) primers were designed to amplify fragments containing the CYP7A1 or CYP7B1 gene exon coding region (6,7). After enzyme processing with ExoSAP-IT (USB, Cleveland, OH), direct sequencing of the amplified PCR products was carried out with the DTCS Quick Start Kit (Beckman Coulter, Fullerton, CA) according to the manufacturer's protocol, using the same primers as for PCR amplification. The sequencing reaction product was analyzed electrophoretically, using a SEQ2000XL analyzer (Beckman Coulter). Digestion of 5 μL of amplified CYP7B1 exon 3 PCR fragment by Taq I (Takara, Shiga, Japan), which recognizes the sequence TCGA, was carried out in a 20-μL reaction buffer at 65°C for 1 hour. A 1.5% agarose gel was used for separation, followed by staining with ethidium bromide. RESULTS Biochemical Identification of an Inborn Error in Bile Acid Synthesis (Table 1) When we analyzed the bile acids in the patient's serum and urine during UDCA treatment at 5 months of age, we detected large amounts of an unusual bile acid, 3β-hydroxy-5-cholen-24-oic acid. However, we did not detect any other unusual bile acid, such as 3β-dihydroxy- and 3β-trihydroxy-Δ5 bile acids and 3-oxo-Δ4 bile acids. The usual bile acids (eg, cholic, chenodeoxycholic, deoxycholic, lithocholic acids) were detected in small amounts or not detected in the serum and urine. The percentages of 3β-hydroxy-5-cholen-24-oic acid relative to total bile acids apart from UDCA were 86.4% and 87.0% in the serum and urine, respectively. Given that we did not have a standard sample of 3β-hydroxy-5-cholesten-27-oic acid, we could not perform that determination.TABLE 1: Bile acid analysis during ursodeoxycholic acid (UDCA) treatment of a patient with oxysterol 7α-hydroxylase deficiencyIdentification of an Oxysterol 7α-Hydroxylase Defect A single substitution of C to T at nucleotide position 538 was confirmed at exon 3 of the CYP7B1 gene, causing an amino acid transition from arginine (CGA) to a stop codon (TGA) at amino acid position 112 (R112→Stop; Fig. 2). The patient was homozygous for this mutation, whereas his parents were heterozygous (Fig. 3).FIG. 2: Pedigree and genomic DNA sequence of the oxysterol 7α-hydroxylase gene in the patient, his parents, and a control. The position of the mutant nucleotide sequence is shown with arrows, indicating C/T in the parents, T in the patient, and C in the control subject. The reverse strand sequence showed the same result. This represents a CGA-to-TGA mutation, affecting arginine at position 112 (R112→Stop), which is replaced by a stop codon. Such a nucleotide substitution was not observed in 102 controls. (The wild-type nucleotide “C” appears in a selected control.)FIG. 3: Digestion of amplified CYP7B1 exon3 polymerase chain reaction (PCR) fragment with Taq I. To screen for the novel C-to-T mutation at nucleotide 538, we amplified the PCR products of the CYP7B1 gene exon 3. PCR products from the normal control allele were digested by the enzyme Taq I into 2 fragments (break point: T/CGA), whereas products with the mutated allele were not digested. The patient was confirmed to be homozygous for the C-to-T substitution, because his product was not digested. Both parents showed digested and nondigested fragments, indicating heterozygosity for the mutation. bp = base pairs.DISCUSSION Generally, the clinical, biochemical, and histological characteristics of inborn errors of bile acid synthesis, such as 3β-hydroxy-C27-steroid dehydrogenase/isomerase deficiency, include prolonged jaundice after birth; a family history of neonatal cholestasis; hepatomegaly; absence of pruritus despite the presence of conjugated hyperbilirubinemia; absence of, or normal concentrations of, total bile acids in serum by an enzymic technique using 3α-hydroxysteroid dehydrogenase; normal serum GGT activity; low serum total cholesterol concentration; and progressive intrahepatic cholestasis with liver fibrosis (8,9). Our patient had these findings, except that his family history was negative and the total cholesterol concentration in his serum was not low. We therefore analyzed bile acids in his serum and urine using GC-MS. These bile acid profiles, such as high concentrations of 3β-monohydroxy-Δ5 bile acid in the serum and urine, closely resembled the findings previously reported in oxysterol 7α-hydroxylase deficiency (1). We have considerable experience with progressive familial intrahepatic cholestasis type 1 or 2, Alagille syndrome, and biliary atresia without pruritus in early infancy. In these cases, to eliminate the diagnosis of cholestasis, we used the following: serum GGT and total bile acids concentrations, liver histology, bile acid analysis, and characteristic facies. In knockout mice lacking a functional cholesterol 7α-hydroxylase enzyme, cholestasis develops before upregulation of oxystrol 7α-hydroxylase activity in the acidic pathway (10–12). Oxysterol 7α-hydroxylase activity first was detected in 3- to 4-week-old mice and remained detectable in the livers of older mice (8). Accordingly, we sought to find mutations of the cholesterol 7α-hydroxylase gene (CYP7A1) in our patient, but no CYP7A1 mutation was present. Instead, we diagnosed the patient as having oxysterol 7α-hydroxylase deficiency after determining that he was homozygous for a mutation in the oxysterol 7α-hydroxylase gene (CYP7B1) by sequence analysis of genomic DNA (Fig. 2). His parents are in excellent health with no liver dysfunction. This finding also suggests that this enzyme deficiency has an autosomal recessive transmission pattern and can be detected by antenatal molecular diagnosis. CYP7B1 is located on chromosome 8q21.3. Although cholesterol 7α-hydroxylase enzyme activity has been reported to be low or absent in this disease, we suspected that the main pathway of bile acid synthesis in our patient could be the acidic pathway, based on the results of the bile acid analysis of his serum and urine. We speculated that the low or absent activity of cholesterol 7α-hydroxylase enzyme activity in this patient reflected previous observations that physiologically, cholesterol 7α-hydroxylase enzyme activity is low or absent in fetal and neonatal life (13,14), and/or the C27 bile acid detected in this disease (3β-monohydroxy-Δ5-C27 bile acid) (1) is a high-affinity ligand for the farnesoid X receptor (15) (Fig. 4).FIG. 4: The effect of defect of oxysterol 7α-hydroxylase. Reduced synthesis of primary bile acids from cholesterol and increased synthesis of 3β-monohydroxy-Δ5 bile acids are shown in the 2 major pathways. The 3β-monohydroxy-Δ5 bile acid, especially 3β-monohydroxy-Δ5-C27 bile acid, is a high affinity ligand for the farnesoid X receptor; therefore, the activity of cholesterol 7α-hydroxylase was reduced.The progressive liver damage in this disease may be rapidly aggravated by the accumulation of 3β-monohydroxy-Δ5 bile acids; our patient developed hepatic cirrhosis by 5 months of age. Similarly, the first reported patient with this disease had cirrhosis at 3 months, and died of chronic liver failure from severe cholestasis and disseminated lymphoproliferative disease after orthotopic liver transplantation (1). We now believe that the current treatment of choice is orthotopic liver transplantation; medical management with UDCA or primary bile acid treatment has failed consistently. Around 600 samples from cholestatic infants have been analyzed by GC-MS during the last 10 years in our laboratory, and this is the first patient identified with oxysterol 7α-hydroxylase deficiency. Setchell et al (1) reported that the prevalence of this disease was lower than 1 in 2500. Overall during this period, we identified 7 patients with an inborn error in bile acid synthesis; the others included 2 with 3β-hydroxy-Δ5-C27-steroid dehydrogenase/isomerase deficiency (16,17), 2 with 3-oxo-Δ4-steroid 5β-reductase deficiency (not published), and 2 with Zellweger syndrome (18). It is difficult to distinguish between each inborn error in bile acid synthesis without bile acid analysis, even if these patients have met all of the previously described clinical, biochemical, and histological characteristics of inborn errors in bile acid synthesis. Finally, when pediatricians encounter patients with an inborn error of bile acid synthesis that fails to yield a specific diagnosis after routine tests, bile acid analysis using GC-MS should be carried out early. However, negative ion fast atom bombardment-mass spectrometry is better than our GC-MS method for the screening of inborn errors in bile acid synthesis (1,7). If the patient is diagnosed with oxysterol 7α-hydroxylase deficiency, then orthotopic liver transplantation should be pursued urgently.