Inborn errors of bile acid biosynthesis are a rare cause of cholestasis that, if not treated, becomes cirrhosis and end-stage liver disease (1). 3β-Hydroxy-Δ5-C27-steroid dehydrogenase/isomerase (3β-HSD) deficiency is the most common of these diseases and accounts for a few cases of low γ-glutamyl transpeptidase cholestatic disorders (2,3). 3β-HSD is a membrane-bound enzyme of the endoplasmic reticulum that catalyses the oxido-reduction of the 3β-hydroxy moiety of 7α-hydroxycholesterol. The lack of this enzyme causes the accumulation of abnormal metabolites unable to function as normal bile acids and sustain the progression of liver disease towards cirrhosis. The defect is corrected by the administration of primary bile acids such as cholic acid (CA) and chenodeoxycholic acid (CDCA), but not by ursodeoxycholic acid (UDCA) (4). Nevertheless, primary bile acids have been reported to be potentially toxic to the hepatocyte (5), and therefore caution should be used with administration and doses of such substances.
Despite several reports of 3β-HSD deficiency that have been published so far (6–9), few studies have thoroughly evaluated the efficacy of treatment, namely the appropriate drug dosage allowing the suppression of the synthesis of toxic metabolites through negative feedback on the bile acid pathway (10).
We therefore aimed to describe the patients diagnosed and followed up at our institution who presented at different ages and with different clinical features. We aimed to reduce primary bile acid supplements to a minimum dose, but with correction of the condition and avoidance of disease progression by a mixture of UDCA/CDCA available on the Italian market.
PATIENTS AND METHODS
Five patients from 2 families were diagnosed at our centre, a pair of siblings 9 and 5 years old (patients 1 and 2) (family A), and 3 siblings of 9 months, 3 years, and 7 years (patients 3, 4, and 5) (family B), respectively. All of the patients, upon diagnosis, were started on a mixture of 50% UDCA and 50% CDCA, the only product containing CDCA on the Italian market (Bilenor, Schwarz Pharma, Monheim, Germany). CA is not available on the Italian market. Patient 1 received UDCA for 22 months and after diagnosis was started on UDCA/CDCA. The patients thereafter were weaned and maintained on the minimum dose of UDCA/CDCA sufficient to provide negative feedback on 3β,7α-dihydroxy-5-cholenoic acid synthesis.
We monitored liver function tests (LFTs) including total and conjugated bilirubin, aspartate aminotransferases (AST), alanine aminotransferase (ALT), γ-glutamyl transpeptidase (γGT), prothrombin time (PT), partial thromboplastin time (PTT), and levels of urinary 3β,7α-glyco-dihydroxy-5-cholenoic acid (u-3β-D-OH-5C), a metabolite abnormally synthesised in 3β-HSD deficiency, to detect the suppression of the bile acid synthetic pathway. This metabolite is a reliable marker of metabolic derangement in this condition that is easily detected by liquid chromatography-tandem mass spectrometry (LC-MS/MS). A level ≤20 μmol/mmol of creatinine (Cr) was considered consistent with effective suppression of the abnormal pathway. The disease was also assessed by liver biopsy in all but 1 patient, who was diagnosed using neonatal screening. After diagnosis, 2 patients had a follow-up biopsy to detect progression of the disease on treatment. 3β-HSD deficiency was suspected on clinical grounds (cholestasis with low γGT and low serum bile acids) or by family history and diagnosed analysing urine bile acid pattern by LC-MS/MS. Urine samples were collected from the patients and stored at −20°C until analysis was carried out. The concentration of bile acids in the urine was corrected according to the Cr concentration and expressed as micromoles per millimole of Cr. For extraction of bile acids from urine we used “OASIS HLB” Waters solid-phase extraction cartridges (1 mL). The cartridges were prewashed with methanol and distilled water. The urine sample, previously normalised for the Cr concentration, was passed through a cartridge with 50 μL of internal standard (100 μM of CA and CDCA conjugated or not with glycine and taurine, labelled with 4 atoms of deuterium). The bile acids and alcohols were eluted with 400 μL of methanol, taken to dryness by a stream of N2, and then diluted with 100 μL of a solution of 50% acetonitrile and 1% ammonia. The sample was analysed by the mass spectrometer micromass QUATTRO II interfaced with HPLC HT Waters 2790 by electrospray ionisation. The analysis was performed by direct injection (without a column). The negative ion mass spectra of the elutes were recorded in the mass range of 380 to 650 Da, which represents the mass ranges of the different urinary bile acids. Data were elaborated by Quantify software (MassLynx, Waters Corp, Milford, MA) (Fig. 1). Normal levels of u-3β-D-OH-5C (L 20 μmol/mmol Cr) were gained from an in-house control group study of healthy children and cholestatic patients with biliary atresia.
Five patients presenting with giant cell hepatitis, biliary cirrhosis, and cryptogenic cirrhosis (1 each) and picked up by neonatal screening (2 patients) were diagnosed at a median age of 2.5 years (range 0.1–5.5 years). A brief case report is provided for each patient from both families.
Family A, Italian Ancestry
Patient 1, a 9-year-old boy born after 40 weeks' gestation without complications, presented at 2.5 months of age with persistent jaundice, hepatomegaly, pale stools, dark urine, total bilirubin 207 μmol/L, conjugated bilirubin 141 μmol/L, AST 334 U/L, ALT 325 U/L, γGT 43 U/L, and serum bile acids 0.5 μmol/L. He was treated with gancyclovir for suspicion of cytomegalovirus hepatitis with no clinical improvement. At 6 months of age, AST was 699 U/L, ALT 567 U/L, total bilirubin 210 μmol/L, and γGT always normal. Following referral to our unit, further investigations showed a normal serum α-1-antitrypsin concentration and genotype, a viral aetiology was ruled out, and cortisol and thyroid function were unremarkable. At 7 months, a liver biopsy demonstrated severe cholestasis and giant cell transformation with progression to porto-portal fibrosis. The child was started on UDCA and fat-soluble vitamins with prompt normalisation of LFT (Fig. 2).
At 2.5 years of age, the stored urine samples collected before starting UDCA were reanalysed retrospectively and the diagnosis of 3β-HSD deficiency was made. The assessment of u-3β-D-OH-5C showed a decrease during UDCA therapy but not the disappearance of this toxic metabolite. A second liver biopsy showed worsened fibrosis despite disappearance of cholestasis and normalisation of LFTs. The child was started on UDCA/CDCA with progressive disappearance of u-3β-D-OH-5C. A follow-up biopsy at 2 years from the start of the treatment showed no progression of fibrosis (Fig. 3). LFTs and physical examination at 9 years of follow-up were normal.
Patient 2, a 6-year-old boy and brother of patient 1, was healthy at birth. Considering the family history, we performed urinanalysis by LC-MS/MS soon after birth, which revealed high concentrations of u-3β-D-OH-5C. LFTs were normal. Serum bile acids were <0.5 μmol/L. The child was started on UDCA/CDCA and followed up. A liver biopsy at 4 years from diagnosis showed minimal portal inflammation with no fibrosis. At 5 years of follow-up, LFTs and physical examination were normal, with normal u-3β-D-OH-5C.
Family B, Moroccan Ancestry
Patient 3, a 24-month-old girl was diagnosed at birth by LC-MS/MS because of the positive family history for 3β-HSD deficiency. LFTs were normal at birth and serum bile acids were <0.5 μmol/L. Started on UDCA/CDCA at 1 month of age, after 2 years' follow-up her levels of u-3β-D-OH-5C were normal, international normalised ratio (INR) 1.1, AST 50 U/L, ALT 21 U/L, γGT 6 U/L, total bilirubin 3 μmol/L, total protein 67 g/L, albumin 42 g/L, and serum bile acids 2 μmol/L.
Patient 4 was 6-year-old boy with a history of failure to thrive, transitory jaundice with pale stools, mucosal bleeding, and hepatosplenomegaly during infancy. He presented to our centre at 3 years of age with severe cholestatic liver disease, normal γGT, high transaminases (ALT 329 U/L), jaundice, severe coagulopathy despite vitamin K supplement (INR 2.3), hepatosplenomegaly (liver at 4 cm from costal border, spleen at 8 cm), and ascites. Serum bile acids were <0.5 μmol/L. A surgical liver biopsy showed severe cholestasis and cirrhosis with moderate inflammation. Urin analysis by LC-MS/MS revealed high concentrations of u-3β-D-OH-5C. He was then started on UDCA/CDCA, with rapid normalisation of urine bile acid pattern. Evaluation at 3 years' follow-up shows INR 1.18, AST 47 U/L, ALT 40 U/L, γGT 8 U/L, total bilirubin 12.5 μmol/L, total protein 72 g/L, albumin 41 g/L, serum bile acid 1.8 μmol/L, and normal u-3β-D-OH-5C. Physical examination shows hepatosplenomegaly with no ascites.
Patient 5 was an 8-year-old girl with a history of jaundice, hepatomegaly, bleeding, and oedema of the legs in the first year of life. She presented to us at 5.5 years of age with normal bilirubin, normal AST and ALT, serum bile acids 1 μmol/L, low γGT, and mild coagulopathy (INR 1.4). The liver biopsy showed liver cirrhosis with minimal inflammation and no cholestasis. Urin analysis by LC-MS/MS revealed high concentrations of u-3β-D-OH-5C. UDCA/CDCA therapy was started with suppression of anomalous bile acids synthesis. LFTs at 3 years follow-up were: INR 1.09, AST 35 U/L, ALT 24 U/L, γGT 8 U/L, total bilirubin 8.7 μmol/L, total protein 71 g/L, albumin 43 g/L, and serum bile acid and 0.8 μmol/L. Urinary 3β-D-OH-5C was normal. Physical examination was unremarkable.
Titration of Bile Acid Supplements
Normal levels of u-3β-D-OH-5C were achieved after a median of 4 (range 3–28) months from the start of treatment (Fig. 4). The minimum dose of UDCA/CDCA required to maintain normal u-3β-D-OH-5C levels was 5/5 mg · kg−1 · day−1.
Patient 1 received a starting dose of 5/5 mg · kg−1 · day−1 at 2.5 years of age. Subsequently, the patient outgrew the dose down to 3/3 mg · kg−1 · day−1 and required dose adjustment to maintain good metabolic control. The metabolite level increased progressively, and therefore we adjusted the dose before it went above 20 μmol/mmol of Cr (Fig. 4, panel A).
Patient 2 was started soon after birth on 7.5/7.5 mg · kg−1 · day−1 of UDCA/CDCA, but because of rapid weight gain, he required a dose adjustment after 2 months to achieve good control of u-3β-D-OH-5C. Eventually, u-3β-D-OH-5C levels remained normal even when the patient outgrew the dose to 5/5 mg · kg−1 · day−1, on which he was maintained thereafter (Fig. 4, panel B).
Patient 3 was diagnosed soon after birth and started on 10/10 mg · kg−1 · day−1 of UDCA/CDCA with rapid decrease of u-3β-D-OH-5C. The child maintained good metabolic control even after she outgrew the initial dose down to 5/5 mg · kg−1 · day−1 (Fig. 4, panel C).
Patient 4 was started on 5/5 mg · kg−1 · day−1 of UDCA/CDCA at 3 years of age with rapid decrease of u-3β-D-OH-5C. After 15 months of follow-up and normal weight gain with consequent dose outgrowth, u-3β-D-OH-5C started increasing and therefore the patient required a dose increase to maintain 5/5 mg · kg−1 · day−1 of the mixture (Fig. 4, panel D).
Patient 5, similar to patient 4, was started on 5/5 mg · kg−1 · day−1 of UDCA/CDCA at 5 years of age with rapid decrease of u-3β-D-OH-5C. After 15 months of follow-up and normal weight gain, a slight increase in u-3β-D-OH-5C guided a dose adjustment to maintain 5/5 mg · kg−1 · day−1, on which she was maintained thereafter (Fig. 4, panel E).
A follow-up biopsy in patients 1 and 2 showed no progression of liver disease, demonstrating that the administered dose of primary bile acid corrected the defect.
3β-HSD deficiency is characterised by the accumulation in blood and urine of abnormal bile acid precursors (di- and trihydroxy-Δ-5-cholenoic acids) that are responsible for liver injury. This condition should be suspected when cholestatic liver disease is associated with a normal serum γGT level, a normal serum bile acids concentration measured by the usual methods, and an absence of pruritus. In this condition cholestasis seems to be mainly due to the lack of the choleretic effect of primary bile acids, whereas progression to cirrhosis may be due to the toxic metabolites accumulating in the liver (11). This disease was once diagnosed by fast atom bombardment ionisation mass spectrometry (12). The diagnosis is now preferably made by LC-MS/MS with demonstration of high levels of urinary di- and trihydroxy-Δ-5-cholenoic acids (13–15). The disease has also been characterised as a recessive autosomal condition caused by the mutation in the HSD3B7 gene (9).
3β-HSD deficiency can present at different ages with different clinical patterns and severity of liver disease. Sometimes the diagnosis can be challenging: The typical clinicopathological pattern is that of neonatal conjugated jaundice with giant cell hepatitis (16). Daugherty et al (17) have shown that, in a different disorder of bile acid metabolism, the histology may improve remarkably after treatment with primary bile acids, despite some residual fibrosis. The follow-up biopsy of patient 1 of our study showed that even in 3β-HSD deficiency this may be the case. Instead, if this disease is left untreated, it will lead to cirrhosis and end-stage liver disease.
We have described here 5 cases, 1 presenting as neonatal giant cell hepatitis, 2 discovered by family screening, 1 presenting as biliary cirrhosis, and 1 having the clinical and histological features of cryptogenic cirrhosis. Whereas the diagnosis can be rather straightforward in children presenting with neonatal cholestatic disease, low γGT, and low serum bile acids, it may not be so in older patients in whom the disease can progress to biliary cirrhosis such as patient 4. In such a clinical picture, cholestasis with low γGT should prompt the differential diagnosis with inherited disorders of bile acid transport such as progressive familial intrahepatic cholestasis type 1 (ATP8B1 disease, PFIC 1) and bile salt export pump deficiency (ABCB11 disease, PFIC 2) (18). Even fewer clues to the diagnosis are provided in cases presenting with noncholestatic cirrhosis, such as patient 5 of our series. In these patients only the bile acid urinary profile performed by LC-MS/MS can warrant the diagnosis.
It has been suggested that UDCA is insufficient to treat this condition, which can be corrected only by the administration of primary bile acid supplements such as CA and CDCA (3,4). Because of the unavailability of such primary bile acids we used a mixture of UDCA/CDCA, the only product available on the Italian market. The strict monitoring we performed with our patients allowed us to gather further information on the effect of bile acid supplements in this setting. Despite that the disease progression can be covered by UDCA monotherapy, which leads to the complete clearance of clinical cholestasis with little effect on toxic metabolites, this bile acid does not seem to be detrimental in itself. In patient 1 of our series, UDCA was efficient in clearing the adverse effects of cholestasis and did not prevent the diagnosis by LC-MS/MS. Nevertheless, UDCA does not correct the metabolic derangement because it does not inhibit cholesterol 7α-hydroxylase and should not be used on its own to treat this disease (16). In fact, in patient 1 abnormal urine metabolites decreased but did not disappear during UDCA therapy (Fig. 4). Although in this patient the appropriate treatment start was followed by no disease progression at histology, this should be interpreted with caution because it is difficult to determine whether it was because of the suppression of the abnormal metabolites or due to the natural history of such a variable disease.
It has been shown that in δ4-3-oxosteroid 5β-reductase deficiency the efficacy of CA in reducing the proportions of the atypical bile acid metabolites is greatest when CA is given alone, rather than in combination with CDCAs or UDCAs (17). This has not been demonstrated in 3β-HSD deficiency. Because CA and CDCA are not available on the Italian market, we decided to study the efficacy of the formulation containing UDCA and CDCA that we have used in the last 15 years for this disease. Clayton et al (19) has demonstrated that a combination of UDCA/CA is effective in controlling 3β-HSD deficiency. We believe our results with UDCA/CDCA can add important information on what has already been demonstrated in 3β-HSD deficiency treated with CA monotherapy or UDCA/CA, especially with regard to monitoring the suppression of the abnormal metabolic pathway.
Despite our achieving good control of this disease by a mixture of UDCA/CDCA, we believe the optimum treatment for this condition is likely CA or CDCA monotherapy, being effective and probably better absorbed if given alone. Because CDCA is going to be available use in our patients in the near future, it will be interesting to use the data presented in the present report to compare the efficacy of CDCA with that of the mixture of UDCA/CDCA that has been on the market for 20 years. In fact considering that UDCA has some beneficial effect on this condition and here it has been used with CDCA, we do not know whether 5 mg of CDCA monotherapy will be sufficient to achieve metabolic control. For this reason, measuring abnormal metabolites during the follow-up of this condition is greatly advisable in patients treated with CDCA monotherapy.
Although primary bile acids correct the disease entirely, they have been regarded as possibly toxic to the liver in in vitro studies (20). The possibility of bile acid toxicity even in in vivo studies justifies our aim to look for the minimum dose titrated to suppress the metabolic pathway causing liver injury in 3β-HSD deficiency (21,22). For this reason, we have monitored the efficacy of CDCA by u-3β-D-OH-5C, a metabolite marking the activity of the bile acid synthetic pathway.
Using our study we could also describe the course of the disease in patients diagnosed at birth and start immediately on proper treatment. Patient 2, diagnosed by family screening, cleared 3β-D-OH-5C urinary levels after administration of UDCA/CDCA and, remarkably, after 4 years of treatment had a liver histology with no fibrosis, suggesting that the defect was corrected.
In conclusion, we have shown that the pattern of presentation of 3β-HSD deficiency can vary from neonatal hepatitis to cryptogenic cirrhosis presenting later in life and whose diagnosis requires a high index of suspicion. Prompt identification of the defect is important to prevent progression to cirrhosis. Patients under titrated doses of a mixture of UDCA/CDCA achieve complete control of the metabolic pathway with complete disappearance of toxic metabolites and histological remission.
1. Bove KE, Heubi JE, Balistreri WF, et al
. Bile acid synthetic defects and liver disease: a comprehensive review. Pediatr Dev Pathol 2004; 7:315–334.
2. Clayton PT, Leonard JV, Lawson AM, et al
. Familial giant cell hepatitis associated with synthesis of 3 beta, 7 alpha-dihydroxy- and 3 beta,7 alpha, 12 alpha-trihydroxy-5-cholenoic acids. J Clin Invest 1987; 79:1031–1038.
3. Heubi JE, Setchell KDR, Bove KE. Inborn errors of bile acid metabolism. Semin Liver Dis 2007; 27:282–294.
4. Yamato Y, Kimura A, Murai T, et al
. 3beta-hydroxy-delta5 -C27-steroid dehydrogenase deficiency: diagnosis and treatment. J Paediatr Child Health 2001; 37:516–519.
5. Sokol RJ, Devereaux M, Dahl R, et al
. “Let there be bile”–understanding hepatic injury in cholestasis. J Pediatr Gastroenterol Nutr 2006; 43:S4–S9.
6. Jacquemin E, Setchell KD, O'Connell NC, et al
. A new cause of progressive intrahepatic cholestasis: 3 beta-hydroxy-C27-steroid dehydrogenase/isomerase deficiency. J Pediatr 1994; 125:379–384.
7. Akobeng AK, Clayton PT, Miller V, et al
. An inborn error of bile acid synthesis (3beta-hydroxy-delta5-C27-steroid dehydrogenase deficiency) presenting as malabsorption leading to rickets. Arch Dis Child 1999; 80:463–465.
8. Kobayashi M, Koike M, Sakiyama M, et al
. 3beta-hydroxy-delta5-C27-steroid dehydrogenase/isomerase deficiency in a 23-year-old woman. Pediatr Int 2000; 42:685–688.
9. Cheng JB, Jacquemin E, Gerhardt M, et al
. Molecular genetics of 3beta-hydroxy-Delta5-C27-steroid oxidoreductase deficiency in 16 patients with loss of bile acid synthesis and liver disease. J Clin Endocrinol Metab 2003; 88:1833–1841.
10. Setchell KDR, Heubi JE. Defects in bile acid biosynthesis-diagnosis and treatment. J Pediatr Gastroenterol Nutr 2006; 43(Suppl 1):S17–S22.
11. Setchell KDR, O'Connell NC. Disorders of bile acid synthesis and metabolism. In: Suchy FJ, Sokol RJ, Balistreri WF, editors. Liver Disease in Children. 3rd ed. New York: Cambridge University Press; 2007. pp. 736–766.
12. Lawson AM, Setchell KDR. Mass spectrometry of bile acids. In: Setchell KDR, Kritchevsky D, Nair PP, editors. The Bile Acids. Volume 4: Methods and Applications. New York: Plenum Press; 1988. pp. 167–268.
13. Mushtaq I, Logan S, Morris M, et al
. Screening of newborn infants for cholestatic hepatobiliary disease with tandem mass spectrometry. BMJ 1999; 319:471–477.
14. Yousef IM, Perwaiz S, Lamireau T, et al
. Urinary bile acid profile in children with inborn errors of bile acid metabolism and chronic cholestasis; screening technique using electrospray tandem mass-spectrometry (ES/MS/MS). Med Sci Monit 2003; 9:MT21–MT31.
15. Perwaiz S, Tuchweber B, Mignault D, et al
. Determination of bile acids in biological fluids by liquid chromatography-electrospray tandem mass spectrometry. J Lipid Res 2001; 42:114–119.
16. Setchell KDR. Disorders of bile acid synthesis. In: Walker WA, Kleinman RE, Sherman PM, et al., editors. Pediatric Gastrointestinal Disease: Pathophysiology, Diagnosis, Management. Toronto: BC Decker Inc; 2004. pp. 1308–1343.
17. Daugherty CC, Setchell KD, Heubi JE, et al
. Resolution of liver biopsy alterations in three siblings with bile acid treatment of an inborn error of bile acid metabolism (delta 4-3-oxosteroid 5 beta-reductase deficiency). Hepatology 1993; 18:1096–1101.
18. Thompson RJ, Strautnieks S. Inherited disorders of transport in the liver. Curr Opin Genet Dev 2000; 10:310–313.
19. Clayton PT, Mills KA, Johnson AW, et al
. Delta 4-3-oxosteroid 5 beta-reductase deficiency: failure of ursodeoxycholic acid treatment and response to chenodeoxycholic acid plus cholic acid. Gut 1996; 38:623–628.
20. Sokol RJ, Dahl R, Devereaux MW, et al
. Human hepatic mitochondria generate reactive oxygen species and undergo the permeability transition in response to hydrophobic bile acids. J Pediatr Gastroenterol Nutr 2005; 41:235–243.
21. Ichimiya H, Nazer H, Gunasekaran T, et al
. Treatment of chronic liver disease caused by 3 beta-hydroxy-delta 5-C27-steroid dehydrogenase deficiency with chenodeoxycholic acid. Arch Dis Child 1990; 65:1121–1124.
22. Ichimiya H, Egestad B, Nazer H, et al
. Bile acids and bile alcohols in a child with hepatic 3 beta-hydroxy-delta 5-C27-steroid dehydrogenase deficiency: effects of chenodeoxycholic acid treatment. J Lipid Res 1991; 32:829–841.