Bile acids for viral hepatitis

Wendong Chen; Jian Ping Liu; Christian Gluud

doi:10.1002/14651858.CD003181.pub2

. 2007 Oct 17;2007(4):CD003181. doi: 10.1002/14651858.CD003181.pub2

Bile acids for viral hepatitis

Wendong Chen ^1,^✉, Jian Ping Liu ², Christian Gluud ³

Editor: Cochrane Hepato‐Biliary Group

PMCID: PMC8711080 PMID: 17943781

Abstract

Background

Trials have assessed bile acids for patients with viral hepatitis, but no consensus has been reached regarding their usefulness.

Objectives

To assess the beneficial and harmful effects of bile acids for viral hepatitis.

Search methods

Searches were performed in The Cochrane Hepato‐Biliary Group Controlled Trials Register (July 2007), The Cochrane Library (Issue 1, 2007), MEDLINE (July 2007), EMBASE (July 2007), Science Citation Index Expanded (July 2007), and Chinese Biomedical Database (July 2007).

Selection criteria

Randomised clinical trials comparing any dose or duration of bile acids versus placebo or no intervention for viral hepatitis were included, irrespective of language, publication status, or blinding. Co‐interventions were allowed in the included randomised clinical trials.

Data collection and analysis

Two authors extracted the data independently. The methodological quality of the trials was evaluated with respect to generation of the allocation sequence, allocation concealment, double blinding, and follow‐up. The outcomes were presented as relative risks (RR) or weighted mean differences (WMD) with 95% confidence intervals (CI).

Main results

We identified 29 randomised trials of bile acids for hepatitis B or C; none were of high methodological quality. We were unable to extract data from two trials. In one trial, ursodeoxycholic acid (UDCA) versus placebo for acute hepatitis B significantly reduced the risk of hepatitis B surface antigen positivity at the end of treatment and serum HBV DNA level at the end of follow‐up. In another trial, UDCA versus no intervention for chronic hepatitis B significantly reduced the risk of having abnormal serum transaminase activities at the end of treatment. Twenty‐five trials compared bile acids (21 trials UDCA; four trials tauro‐UDCA) versus placebo or no intervention with or without co‐interventions for chronic hepatitis C. Bile acids did not significantly reduce the risk of having detectable serum HCV RNA (RR 0.99, 95% CI 0.91 to 1.07), cirrhosis, or portal and periportal inflammation score at the end of treatment. Bile acids significantly decreased the risk of having abnormal serum alanine aminotransferase activity at the end of treatment (RR 0.82, 95% CI 0.76 to 0.90) and follow‐up (RR 0.91, 95% CI 0.85 to 0.98). Bile acids significantly increased the Knodell score (WMD 0.20, 95% CI 0.08 to 0.31) at the end of treatment. No severe adverse events were reported. We did not identify trials including patients with hepatitis A, acute hepatitis C, hepatitis D, or hepatitis E.

Authors' conclusions

Bile acids lead to a significant improvement in serum transaminase activities in hepatitis B and C but have no effects on the clearance of virus. There is insufficient evidence either to support or to refute effects on long‐term outcomes including hepatocellular carcinoma, hepatic decompensation, and liver related mortality. Randomised trials with high methodological quality are required before clinical use is considered.

Plain language summary

Bile acids may improve liver biochemistry of patients with hepatitis B or C, but there is insufficient evidence about long‐term beneficial effects

Viral hepatitis causes significant morbidity and mortality. Based on this Cochrane systematic review, bile acids may decrease serum transaminase activities in patients with acute hepatitis B, chronic hepatitis B, or chronic hepatitis C. However, bile acids have no effects in eradicating viral markers. There is insufficient evidence either to support or to refute effects on the long‐term outcomes that include hepatocellular carcinoma, decompensated cirrhosis, and/or liver related mortality.

No clinical trials have evaluated bile acids for patients with hepatitis A, acute hepatitis C, hepatitis D, or hepatitis E. Accordingly, bile acids should first be evaluated in adequately sized and adequately conducted randomised trials before clinical use is considered.

Background

Viral hepatitis is liver inflammation that can be caused by hepatitis A, B, C, D, or E virus (Wolfram 1999). All of these viruses can cause an acute disease lasting several weeks including jaundice, dark urine, extreme fatigue, nausea, vomiting, and abdominal pain (Zuckermann 1993). After acute hepatitis, Hepatitis B virus (HBV), hepatitis C virus (HCV), and hepatitis D virus (which can only infect patients already infected by hepatitis B virus) also cause chronic infection, which may develop into hepatic cirrhosis and/or liver cancer (Rizzetto 1986; Fattovich 1995; Brechot 1996). Viral hepatitis is considered prevalent disease associated with significant morbidity, mortality, and socio‐economic loss (Stapleton 1994). It has been estimated that more than 350 million have chronic HBV infection (WHO 2000a) and 170 million persons are chronically infected with HCV (WHO 2000b).

Current treatment of acute viral hepatitis is based on supportive measures, relief of symptoms, and avoidance of further liver injury (Maddrey 1980). At present, antiviral agents are not advocated for patients with typical acute viral hepatitis (Alam 1995) apart from the use of interferon for acute hepatitis C (Myers 2001a; Seeff 2002). For chronic viral hepatitis, treatment recommendations vary with the viruses. Interferon or lamivudine are used for chronic hepatitis B (Liaw 1999; Lok 2001). Combination therapy with interferon and ribavirin is recommended for patients with chronic hepatitis C (Farrell 1999; Seeff 2002; Brok 2005). Interferon can induce good biochemical responses in patients with chronic hepatitis D, but in the great majority of responders there is relapse after discontinuation of the drug (Wolfram 1999). Interferon exerts a direct antiviral effect and enhances the immunomodulatory response of the host (Hoffnagle 1997; Poynard 1997; Farrell 1999; Liaw 1999). Unfortunately, only a limited number of patients with chronic viral hepatitis are suitable for interferon therapy (Myers 2002b). Even the most effective therapy available, pegylated interferon alpha and ribavirin, leads to a sustained virologic response in about 50% of patients with genotype 1 virus, the major genotype for chronic HCV in western countries (Manns 2001; Fried 2002). New options to treat viral hepatitis have been explored and bile acids have been investigated for the possible effects on viral hepatitis. Moreover, treatment is costly and entails a high rate of adverse events (Renault 1989; Brok 2005). Effective, well tolerated, and economic drugs for the treatment of viral hepatitis are needed.     Bile acids include chenodeoxycholic acid, deoxycholic acid, lithocholic acid, and ursodeoxycholic acid (UDCA) (Fuchs 1999). In the course of cholestasis, intrahepatic accumulation of hydrophobic bile acids such as chenodeoxycholic acid and deoxycholic acid is thought to induce liver damage (Palmer 1972) due to their detergent activity, which appears to dissolve phospholipid and cholesterol in cell membranes (Sholmerich 1984). The hepatocyte toxicity of hydrophobic bile acids has been observed both in animals and human beings (Jaeschke 2002). These effects can be prevented by the addition of UDCA, which modifies the bile acid pool, resulting in an increase of the hydrophilic fraction, and stabilises cell membranes (Armstrong 1982). The mechanisms of UDCA action include reduction of toxic endogenous bile acids, membrane stabilising activity, and an immunomodulatory effect (Calmus 1990). Another bile acid, tauro‐ursodeoxycholic acid (TUDCA), has also been used for 'liver protection' (Podda 1996).

UDCA and TUDCA have been used for the treatment of cholestatic liver diseases including primary biliary cirrhosis (Gluud 2001b), primary sclerosis cholangitis (Chen 2002), and other chronic cholestatic diseases (Larghi 1997). Several randomised clinical trials have shown that UDCA improves the serum biochemical indexes of cholestasis and cytolysis in patients with chronic liver diseases (Nakagawa 1990; Gluud 2001b). Moreover, there is some evidence that UDCA, used alone or in combination with interferon, is able to improve liver enzyme levels in patients with chronic hepatitis C (Boucher 1995). TUDCA has physiochemical and metabolic properties, which favour its use as an alternative to UDCA for chronic cholestatic liver disease (Gallo 1993). TUDCA may prove of benefit also for necroinflammatory liver diseases, especially for HCV‐related chronic hepatitis in which bile duct damage is frequently seen (Fallon 1997).

This Cochrane systematic review represents an update of our previous review, in which we found moderate beneficial effects of bile acids for viral hepatitis in trials with high risk of bias (Chen 2003).

Objectives

The objectives were to evaluate the beneficial and harmful effects of bile acids for viral hepatitis caused by one or more of the hepatotrophic viruses (A to E) based on the results of randomised clinical trials.

Methods

Criteria for considering studies for this review

Types of studies

Randomised clinical trials were included irrespective of blinding, publication status, or language. Quasi‐randomised trials (eg, using date of birth) and observational studies were excluded. However, for the evaluation of rare, serious adverse events, we also considered large cohort studies and previous meta‐analyses/systematic reviews, as such events are rarely captured in randomised clinical trials due to their sample size.

Types of participants

Patients with viral hepatitis diagnosed by any method according to the following definitions were included:   (1) Acute hepatitis A defined as: serum transaminase activities (alanine aminotransferase (ALT); aspartate aminotransferase (AST); gamma glutamyltranspeptidase (GGT)) above the upper normal limits; seropositivity for immunoglobulin M (IgM) anti‐hepatitis A virus.   (2) Acute hepatitis B defined as: serum transaminase activities above twice the upper normal limits; seropositivity for hepatitis B surface antigen (HBsAg), hepatitis B 'e' antigen (HBeAg), and IgM antibody to hepatitis B core antigen.   (3) Chronic hepatitis B defined as: serum transaminase activities above the upper normal limits; serum HBsAg positivity and HBeAg positivity or HBV DNA positivity for more than six months.   (4) Acute hepatitis C defined as: serum transaminase activities above twice the upper normal limits and serum antibody to HCV (anti‐HCV) and/or serum HCV RNA positivity.   (5) Chronic hepatitis C defined as: serum transaminase activities above the upper normal limits for more than six months and/or liver biopsy findings compatible with chronic hepatitis C plus HCV‐RNA seropositivity for more than six months.   (6) Acute hepatitis D defined as: serum transaminase activities above the upper normal limits and seropositivity for hepatitis D virus RNA and IgM anti‐hepatitis D virus.   (7) Chronic hepatitis D defined as: serum transaminase activities above the upper normal limits and hepatitis D virus RNA positivity for more than six months.   (8) Acute hepatitis E defined as: serum transaminase activities above the upper normal limits and seropositivity for hepatitis E virus RNA and IgM anti‐hepatitis E virus.

Types of interventions

This review includes randomised comparisons of any dose or duration of bile acid (UDCA or TUDCA) versus placebo or no intervention. Co‐interventions were allowed if received by both intervention arms.

Types of outcome measures

For each individual type of viral hepatitis, the primary outcome measures were:   (1) Virological response:

acute hepatitis A: detectable hepatitis A virus at the end of treatment and at the maximal follow‐up;
acute hepatitis B: detectable HBsAg, HBV DNA measured by polymerase chain reaction (PCR) technique, HBeAg and seroconversion from HBeAg to anti‐HBe at the end of treatment and at the maximal follow‐up;
chronic hepatitis B: detectable HBsAg, HBV DNA measured by PCR technique, HBeAg and seroconversion from HBeAg to anti‐HBe at the end of treatment and at the maximal follow‐up;
acute hepatitis C: detectable HCV RNA measured by a validated PCR technique at the end of treatment and at the maximal follow‐up;
chronic hepatitis C: detectable HCV RNA measured by a validated PCR technique at the end of treatment and at the maximal follow‐up;
acute hepatitis D: detectable hepatitis D virus RNA measured by a validated PCR technique at the end of treatment and at the maximal follow‐up;
chronic hepatitis D: detectable hepatitis D virus RNA measured by a validated PCR technique at the end of treatment and at the maximal follow‐up;
acute hepatitis E: detectable hepatitis E virus RNA measured by a validated PCR technique at the end of treatment and at the maximal follow‐up.

(2) All‐cause mortality.   (3) Liver‐related morbidity: decompensated cirrhosis, hepatocellular carcinoma, or liver transplantation.

The secondary outcome measures were:   (1) Biochemical non‐responders: number of patients without normalisation of serum transaminase activities at the end of treatment and/or maximal follow‐up.   (2) Histological response: histological activity index or fibrosis score at the end of treatment and/or the maximal follow‐up.   (3) Quality of life.   (4) Adverse events. Number and type of adverse events, defined as any untoward medical occurrence in a patient. This occurrence should not necessarily have a causal relationship with the treatment, but should have resulted in the discontinuation of treatment. We have defined serious adverse events according to the International Conference on Harmonisation (ICH) Guidelines (ICH‐GCP 1997) as any event that leads to death, is life‐threatening, requires in‐patient hospitalisation or prolongation of existing hospitalisation, results in persistent or significant disability, and any important medical event which may have jeopardised the patient or requires intervention to prevent it. All other adverse events were considered non‐serious.   (5) Cost‐effectiveness.

Search methods for identification of studies

The Cochrane Hepato‐Biliary Group Controlled Trials Register (July 2007) and The Cochrane Controlled Trials Register (Cochrane Library, Issue 1, 2007) were searched using the terms 'hepatitis' and the names of individual bile acids (lithocholic acid, or chenodeoxycholic acid, or ursodeoxycholic acid, or deoxycholic acid, or dehydrocholic acid, or tauro‐ursodeoxycholic acid). MEDLINE (1950 to July 2007) was searched using the search strategy of The Cochrane Hepato‐Biliary Group (see Collaborative Review Group for details) combined with the terms 'hepatitis' and the bile acids mentioned above. EMBASE (January 1980 ‐ July 2007) and Science Citation Index Expanded (1945 ‐ July 2007) were searched using the same search terms. In addition, Chinese Biomedical Database (1980 ‐ 2007) was searched by using similar search strategy (Appendix 1).

The principal authors of the included trials and pharmaceutical companies involved in the production of bile acids were contacted about additional published or unpublished randomised clinical trials on the topic.

Data collection and analysis

The meta‐analysis was conducted according to our previously published protocol (Chen 2001) following the recommendations given by The Cochrane Reviewers' Handbook (Higgins 2005). Identified trials were listed and two contributors independently evaluated whether the trials fulfilled the inclusion criteria. Excluded trials were listed with the reasons for exclusion.

Methodological quality   Methodological quality was defined as the confidence that the design and report restrict bias in the intervention comparison (Moher 1998). The methodological quality was assessed by quality components, ie, adequacy of generation of the allocation sequence, allocation concealment, double blinding, and follow‐up (Schulz 1995; Moher 1998; Kjaergard 2001). Trials were assessed as having high methodological quality, ie, low risk of bias, if all quality components were adequate.

The quality components were assessed according to Kjaergard et al (Kjaergard 2001):

Generation of the allocation sequence

Adequate, if the allocation sequence was generated by a computer or random number table. Drawing of lots, tossing of a coin, shuffling of cards, or throwing dice was considered as adequate if a person who was not otherwise involved in the recruitment of participants performed the procedure.
Unclear, if the trial was described as randomised, but the method used for the allocation sequence generation was not described.
Inadequate, if a system involving dates, names, or admittance numbers were used for the allocation of patients. Such quasi‐randomised studies were excluded.

Allocation concealment

Adequate, if the allocation of patients involved a central independent unit, on‐site locked computer, identically appearing numbered drug bottles or containers prepared by an independent pharmacist or investigator, or serially numbered, sealed, and opaque envelopes.
Unclear, if the trial was described as randomised, but the method used to conceal the allocation was not described.
Inadequate, if the allocation sequence was known to the investigators who assigned participants or if the study was quasi‐randomised. The latter studies were excluded.

Blinding

Adequate, if the trial was described as double blind and the method of blinding involved identical placebo or active drugs.
Unclear, if the trial was described as double blind, but the method of blinding was not described.
Not performed, if the trial was not double blind.

Follow‐up

Adequate, if the numbers and reasons for dropouts and withdrawals in all intervention groups were described or if it was specified that there were no dropouts or withdrawals.
Unclear, if the report gave the impression that there had been no dropouts or withdrawals, but this was not specifically stated.
Inadequate, if the number or reasons for dropouts and withdrawals were not described.

Further, we registered whether or not the trials reported to have used an intention‐to‐treat analysis and performed a sample size calculation (Gluud 2001a).

Data extraction   The data were extracted by WC and JL independently. The data were validated by CG. Disagreements were resolved by discussion. The following characteristics were extracted from each trial: primary author, number of patients randomised, patient inclusion and exclusion criteria, methodological quality, sample size calculation, intention‐to‐treat analysis, intervention regimens, mean age, proportion of males, proportion of patients with cholestasis, proportion of patients with cirrhosis, proportion of virus genotypes, number of outcomes, and number and type of adverse events in the intervention and the control group. Further information was sought by correspondence with the principal investigators of the included trials in case the relevant data were not published.

Statistical methods   The analyses were performed in Review Manager 4.1. The analyses for binary outcomes included all patients irrespective of compliance or follow‐up ('intention‐to‐treat') using the last reported observed response ('carry forward'). The analyses for continuous outcomes included only the patients with available data. We used both random‐effects (DerSimonian 1986) and fixed‐effect models (DeMets 1987) in the meta‐analyses. When we did not find a difference between the results of the two methods, we only reported the results of the fixed‐effect model. Evidence of publication bias was analysed by funnel plot analyses (Egger 1997). Binary outcome measures were analysed by relative risks (RR) with 95% confidence intervals (CI) and continuous outcome measures by weighted mean differences (WMD) and 95% CI.

In our protocol (Chen 2001), we planned to perform sensitivity analyses according to the (1) methodological quality of the included trials (adequate compared to inadequate); (2) dose and duration of treatment with bile acid; (3) treatment with or without co‐intervention; and (4) publication status. The trials included in our review were of low methodological quality; the dosages of UDCA applied in the trials were in the range of 10 to 15 mg/kg/day; and we were unable to identify any unpublished trials. Accordingly, sensitivity analyses were only performed according to the duration of treatment with bile acids. The outcome measures used for sensitivity analyses were the risk of having serum HCV RNA at the end of follow‐up and the risk of abnormal serum ALT at the end of treatment and follow‐up.

We planned to do meta‐regression if a sufficient number of trials were identified to explore the effect of patient characteristics, intervention regimens, and methodological quality of the trials on the virological response at the end of treatment. However, we decided not to perform such analyses due to the general low quality of the included trials in this review.

Results

Description of studies

Our initial searches identified 174 studies; 154 from the electronic searches and 20 from the handsearches. After reading titles and abstracts, 124 of these articles were excluded because they were duplicates, non‐clinical studies, or had study objectives different from our review. A total of 50 articles published in three languages (English, Chinese, and Italian) were retrieved for further assessment. Of these, 17 studies were excluded because they did not meet the inclusion criteria. The reasons for exclusion are listed in the table 'Characteristics of excluded studies'.

The remaining 33 publications were randomised clinical trials that reported the random allocation of patients with viral hepatitis to bile acids versus control, including four double publications (Angelico 1995a; Boucher 1995; Puoti 1995; Fabbri 2000). Accordingly, the searches identified 29 trials that met all pre‐specified inclusion criteria. These trials are listed in the table 'Characteristics of included studies'. Twenty‐four trials were published in English and five in Chinese. Ten trials were from Italy, five from China, five from Japan, four from France, two from Argentina, and one each from the Czech Republic, USA, and Turkey.

We were unable to supplement our electronic searches and handsearches with additional data from personal communications. Only one author replied and this author was not able to supply the information because the original data were lost. No responses were received from pharmaceutical companies (Axcan Pharma Inc and Dr. Falk Pharma).

Patients' origin and sample size   One trial (Galský 1999) included in‐patients, three trials (Puoti 1995; Abdelmalek 1998; Fabbri 2000) included out‐patients, and the other 25 trials did not specify the origin of the patients. The average size of the trial was 69 patients (range 31 to 167 patients).

Diagnoses   Inclusion criteria were explicit viral hepatitis diagnosed by virological markers, clinical manifestations, and biochemical variables. Thirteen trials confirmed the diagnosis by liver biopsy. Galský1999 included patients with acute hepatitis including A, B, and/or C. Zhou 1995 included patients with chronic hepatitis B. Twenty‐five trials included patients with chronic hepatitis C. Two trials (Qing 1999; Cadranel 2003) included patients with chronic hepatitis B or C. We were unable to extract data on the respective diagnostic groups, nor did we receive a reply from the authors when requested for separate data on the patients with hepatitis B and C. Accordingly, outcome data from these two trials could not be included in this review. However, they are listed in the table 'Characteristics of included studies' and the section regarding the methodological quality of the included studies. No trials were found evaluating bile acids for hepatitis A, acute hepatitis C, acute or chronic hepatitis D, or hepatitis E.

Bile acids and collateral interventions   Twenty‐five trials evaluated UDCA, and four trials evaluated TUDCA. The median duration of treatment was nine months (range, three to 18 months). The median duration of follow‐up was nine months (range six to 18 months).

Acute or chronic hepatitis B   One trial (Galský1999) compared UDCA versus placebo for acute hepatitis, which included hepatitis A, B, and/or C. Most of patients included in this trial had acute hepatitis B. The dose of UDCA was 750 mg/day and the duration of treatment was three months. The follow‐up after treatment was nine months.

One trial (Zhou 1995) compared UDCA versus no intervention for chronic hepatitis B. Patients in the UDCA group received 400 mg/day for three months. The follow‐up time was three months.

Chronic hepatitis C   Five trials (Leri 1994; Takano 1994; Zhu 1994; Puoti 1995; Scotto 1997) compared UDCA versus placebo or no intervention for chronic hepatitis C. The dose of UDCA ranged from 400 to 800 mg/day. The duration of treatment ranged from three to 12 months. The duration of follow‐up after treatment ranged from three to six months.

Two trials compared TUDCA versus placebo (Crosignani 1998) or no treatment (Belloni 1999) for chronic hepatitis C. The dose of TUDCA ranged from 500 to 750 mg/day; the duration of treatment ranged from four to six months.

Fifteen trials compared UDCA combined with interferon versus interferon monotherapy in patients with chronic hepatitis C. Eleven trials (Clerici 1994; Kawata 1994; Angelico 1995a; Boucher 1995; Tanaka 1996; Ge 1997; Huang 1997; Kiso 1997; Senturk 1997; Abdelmalek 1998; Viola 1998) included patients who were naive to interferon, and four trials (Bonnand 1996; Fabbri 2000; Poupon 2000; Viola 1996) included patients who were non‐responders to interferon. The dose of UDCA ranged from 10 to 15 mg/kg body weight/day, and the dose of interferon ranged from three to six million units three times a week. The duration of treatment ranged from six to 18 months. The duration of follow‐up after treatment ranged from six to 18 months.

Two trials (Picciotto 1994; Pigozzi 1997) compared TUDCA plus interferon versus interferon monotherapy for chronic hepatitis C. The dose of TUDCA ranged from 500 mg/day to 10 mg/kg body weight/day. The dose of interferon was three million units three times a week. The duration of treatment was six months and the follow‐up after treatment ranged from three to six months.

One trial (Tsubota 1999) compared UDCA combined with glycyrrhizin versus glycyrrhizin monotherapy in patients with chronic hepatitis C. The dose of UDCA was 600 mg/day and the dose of glycyrrhizin was 300 ml/week. The duration of treatment was six months.

Chronic hepatitis B and C   Two trials compared UDCA versus placebo (Cadranel 2003) or no intervention (Qing 1999) for chronic hepatitis (including hepatitis B and C). Patients in the Cadranel 2003 trial received 800 mg/day UDCA for 12 months, and patients in the Qing 1999 trial received 450 mg/day UDCA for three months.

Outcome measures   The outcome measures reported by most trials were virological markers and/or biochemical variables. None of the 29 trials reported mortality, hepatocellular carcinoma, liver transplantation, quality of life, or cost‐effectiveness.

Risk of bias in included studies

Of the 29 included trials, one trials (3.5%) (Angelico 1995a) reported both adequate generation of the allocation sequence and adequate allocation concealment. One trial (Boucher 1995) reported only adequate generation of the allocation sequence, and two trials (Picciotto 1994; Tsubota 1999) reported only adequate allocation concealment. The remaining trials had unclear generation of the allocation sequence and allocation concealment. Eight trials (27.6%) (Bonnand 1996; Cadranel 1996; Viola 1996; Abdelmalek 1998; Crosignani 1998; Viola 1998; Galský 1999; Poupon 2000) used double blinding, which was assessed as adequate. One trial (Scotto 1997) reported single blinding, but it was unclear who was blinded. The remaining trials were conducted without blinding. Accordingly, none of the trials were of high methodological quality, ie, having adequate generation of the allocation sequence, allocation concealment, double blinding, and follow‐up. Therefore, all trials were considered as having risk of bias.

Only one trial reported a sample size calculation (3.4%) (Angelico 1995a), and only eight trials (27.6%) (Picciotto 1994; Angelico 1995a; Boucher 1995; Viola 1996; Kiso 1997; Viola 1998; Poupon 2000; Cadranel 2003) performed intention‐to‐treat analyses. Nine trials (31.0%) (Picciotto 1994; Boucher 1995; Kiso 1997; Senturk 1997; Abdelmalek 1998; Crosignani 1998; Galský 1999; Tsubota 1999; Poupon 2000) reported the numbers of patients withdrawn and reasons for withdrawal.

Of the 29 trials, 17 trials (58.6%) (Clerici 1994; Kawata 1994; Picciotto 1994; Angelico 1995a; Boucher 1995; Zhou 1995; Bonnand 1996; Ge 1997; Huang 1997; Kiso 1997; Gracielle 2002; Scotto 1997; Abdelmalek 1998; Galský 1999; Fabbri 2000; Poupon 2000) reported follow‐up after the end of intervention with bile acids. The duration of follow‐up varied from three to 18 months.

Effects of interventions

Bile acids for acute hepatitis B   One trial (Galský 1999) evaluated UDCA versus placebo in patients with acute hepatitis (12 HAV, 61 HBV, 4 HCV, and 1 HBV + HCV). However, only the outcome measures of patients with acute hepatitis B were reported in this trial. UDCA significantly decreased the risk of having positive serum HBsAg at the end of treatment (6/34 (17.6%) versus 12/27 (44.4%); RR 0.40, 95% CI 0.17 to 0.92) (Comparison 01‐01). A non‐significant tendency favouring UDCA was shown at the end of follow‐up (2/34 (5.9%) versus 7/27 (25.9%); RR 0.23, 95% CI 0.05 to 1.00) (Comparison 01‐02). UDCA did not significantly decrease the serum HBV DNA level at the end of treatment (WMD ‐574 pg/ml, 95% CI ‐1149 to 1) (Comparison 01‐03). At the end of follow‐up, UDCA significantly decreased the serum HBV DNA level (WMD ‐955 pg/ml, 95% CI ‐1814 to ‐96) (Comparison 01‐04).

No data were available regarding clinical outcomes including mortality or histological outcomes.

UDCA significantly decreased the risk of abnormal serum GGT activity (4/34 (11.8%) versus 10/27 (37%); RR 0.32, 95% CI 0.11 to 0.90) (Comparison 01‐07), but not ALT activity (4/34 (11.8%) versus 9/27 (33.3%); RR 0.35, 95% CI 0.12 to 1.02) at the end of treatment (Comparison 01‐05). UDCA did not significantly decrease the risk of abnormal serum ALT (2/34 (5.9%) versus 6/27 (22.2%); RR 0.26, 95% CI 0.06 to 1.21) (Comparison 01‐06) or GGT (2/34 (5.9%) versus 6/27 (22.2%); RR 0.26, 95% CI 0.06 to 1.21) activities at the end of follow‐up (Comparison 01‐08).

Bile acids for chronic hepatitis B   One trial (Zhou 1995) evaluated UDCA versus no intervention in 112 patients with chronic hepatitis B. There were no data on virological, clinical, or histological outcomes. UDCA significantly reduced the risk of having abnormal serum ALT activity (26/64 (40.6%) versus 30/48 (62.5%); RR 0.65, 95% CI 0.45 to 0.94) (Comparison 02‐01) and serum ALT activity (WMD ‐15 IU/L, 95% CI ‐21 to ‐9) at the end of treatment (Comparison 02‐02).

Bile acids for chronic hepatitis C   Risk of being serum HCV RNA positive at the end of treatment   Thirteen trials were included in this comparison. Two trials (106 patients) compared UDCA versus placebo (Scotto 1997) or no intervention (Zhu 1994), and 11 trials (831 patients) compared UDCA plus interferon versus interferon (Kawata 1994; Angelico 1995a; Boucher 1995; Bonnand 1996; Tanaka 1996; Ge 1997; Huang 1997; Abdelmalek 1998; Viola 1998; Fabbri 2000). UDCA did not significantly reduce the risk of having serum HCV RNA at the end of treatment (305/419 (72.8%) versus 307/418 (73.4%); RR 0.99, 95% CI 0.91 to 1.07). The trials comparing UDCA versus placebo or no intervention (53/54 (98.2%) versus 51/52 (98.1%); RR 1.01, 95% CI 0.84 to 1.21) and UDCA plus interferon versus interferon (252/365 (69%) versus 256/366 (70%); RR 0.98, 95% CI 0.90 to 1.08) did not individually show significant effects of UDCA (Comparison 03‐01).

We performed a sensitivity analysis according to the duration of treatment. Eight trials (Zhu 1994; Angelico 1995a; Boucher 1995; Bonnand 1996; Ge 1997; Huang 1997; Scotto 1997; Abdelmalek 1998) evaluated short treatment duration (less than 12 months) and five trials (Kawata 1994; Tanaka 1996; Kiso 1997; Viola 1998; Fabbri 2000) evaluated long treatment duration (12 months or more). In none of the subgroups did UDCA significantly decrease the risk of having serum HCV RNA at the end of treatment (short treatment duration: 170/223 (76.2%) versus 178/221 (80.5%); RR 0.93, 95% CI 0.83 to 1.04 and long treatment duration: 135/196 (68.9%) versus 129/197 (65.5%); RR 1.05, 95% CI 0.92 to 1.19) (Comparison 03‐02).

Risk of being serum HCV RNA positive at the end of follow‐up   Ten trials (Kawata 1994; Angelico 1995a; Boucher 1995; Tanaka 1996; Ge 1997; Kiso 1997; Senturk 1997; Abdelmalek 1998; Viola 1998; Fabbri 2000) compared UDCA plus interferon versus interferon (676 patients). UDCA plus interferon did not significantly decrease the risk of having serum HCV RNA at the end of follow‐up (268/333 (80.5%) versus 294/343 (85.7%); RR 0.93, 95% CI 0.87 to 1.00) (Comparison 03‐03).

We performed a sensitivity analysis regarding treatment duration. Five trials (Angelico 1995a; Boucher 1995; Ge 1997; Senturk 1997; Abdelmalek 1998) evaluated short treatment duration (less than 12 months) and showed that UDCA did not significantly decrease the risk of having serum HCV RNA at the end of follow‐up (116/137 (84.7%) versus 125/146 (85.6%); RR 0.99, 95% CI 0.87 to 1.12). Five trials (Kawata 1994; Tanaka 1996; Kiso 1997; Viola 1998; Fabbri 2000) evaluated long treatment duration (12 months or more) and showed that UDCA significantly decreased the risk of serum HCV RNA at the end of follow‐up (152/196 (77.6%) versus 169/197 (85.8%); RR 0.90, 95% CI 0.83 to 0.99) (Comparison 03‐04).

Mortality   We were not able to find any data related to mortality in the trials included in our review.

Risk of cirrhosis at the end of treatment   Fabbri 2000 compared UDCA plus interferon versus interferon alone. UDCA plus interferon did not significantly reduce the risk of cirrhosis at the end of treatment (19/53 (35.9%) versus 10/50 (20%); RR 1.79, 95% CI 0.93 to 3.47) (Comparison 03‐05).

Risk of abnormal serum ALT activity at the end of treatment   Twenty one trials (1582 patients) were included in this comparison. Bile acids significantly decreased the risk of having abnormal serum ALT activity (425/812 (52.3%) versus 467/770 (60.6%); RR 0.83, 95% CI 0.77 to 0.90) (Comparison 03‐06). Three trials (Takano 1994; Zhu 1994; Puoti 1995) compared UDCA versus placebo or no intervention, 15 trials (Clerici 1994; Kawata 1994; Angelico 1995a; Boucher 1995; Bonnand 1996; Tanaka 1996; Viola 1996; Ge 1997; Huang 1997; Kiso 1997; Senturk 1997; Abdelmalek 1998; Viola 1998; Fabbri 2000; Poupon 2000) compared UDCA plus interferon versus interferon, two trials (Picciotto 1994; Gracielle 2002) compared TUDCA plus interferon versus interferon, and one trial (Tsubota 1999) compared UDCA plus glycyrrhizin versus glycyrrhizin. UDCA significantly decreased the risk of having abnormal serum ALT activity at the end of treatment in the comparisons of UDCA versus placebo or no intervention (97/141 (68.8%) versus 81/98 (82.7%); RR 0.77, 95% CI 0.68 to 0.87) and UDCA plus interferon versus interferon (236/501 (47.1%) versus 291/506 (57.5%); RR 0.81, 95% CI 0.73 to 0.91). Bile acids (UDCA and TUDCA) did not significantly decrease the risk of having abnormal serum ALT activity in the comparison of TUDCA plus interferon versus interferon (43/85 (50.6%) versus 44/81(54.3%); RR 0.93, 95% CI 0.75 to 1.24) and UDCA plus glycyrrhizin versus glycyrrhizin (49/85 (57.7%) versus 51/85 (60%); RR 0.96, 95% CI 0.75 to 1.24) (Comparison 03‐06).

A sensitivity analysis was performed regarding treatment duration. Fifteen trials (Clerici 1994; Picciotto 1994; Takano 1994; Zhu 1994; Angelico 1995a; Boucher 1995; Puoti 1995; Bonnand 1996; Viola 1996; Ge 1997; Huang 1997; Senturk 1997; Abdelmalek 1998; Tsubota 1999; Poupon 2000) showed that short‐term treatment (less than 12 months) with bile acids (UDCA or TUDCA) significantly decreased the risk of having abnormal serum ALT activity at the end of treatment (301/561 (53.7%) versus 324/522 (62.1%); RR 0.82, 95% CI 0.75 to 0.91). Five trials (Kawata 1994; Tanaka 1996; Kiso 1997; Viola 1998; Fabbri 2000) showed that long‐term treatment (12 months or more) with UDCA significantly decreased the risk of having abnormal serum ALT activity at the end of treatment (93/196 (47.4%) versus 113/197 (57.4%); RR 0.83, 95% CI 0.69 to 0.99) (Comparison 03‐07). No significant difference was observed between short‐term treatment and long‐term treatment (P = 0.89).

Risk of abnormal serum ALT at the end of follow‐up   Fourteen trials (997 patients) were included in this comparison. Bile acids (UDCA and TUDCA) significantly decreased the risk of abnormal serum ALT activity at the end of follow‐up (351/500 (70.2%) versus 384/497 (77.3%); RR 0.91, 95% CI 0.85 to 0.98) (Comparison 03‐08). Twelve trials (Clerici 1994; Kawata 1994; Boucher 1995; Bonnand 1996; Tanaka 1996; Ge 1997; Kiso 1997; Senturk 1997; Abdelmalek 1998; Viola 1998; Fabbri 2000; Poupon 2000) compared UDCA plus interferon versus interferon and two trials (Picciotto 1994; Pigozzi 1997) compared TUDCA plus interferon versus interferon. UDCA plus interferon significantly decreased the risk of abnormal serum ALT at the end of follow‐up (317/434 (73%) versus 352/435 (80.9%); RR 0.90, 95% CI 0.84 to 0.97), while TUDCA plus interferon did not show this effect (34/66 (51.5%) versus 32/62 (51.6%); RR 1.01, 95% CI 0.73 to 1.40).

Sensitivity analysis was performed regarding treatment duration. Eight trials (Picciotto 1994; Boucher 1995; Bonnand 1996; Ge 1997; Pigozzi 1997; Senturk 1997; Abdelmalek 1998; Poupon 2000) showed that short‐term treatment (less than 12 months) with bile acids (UDCA or TUDCA) did not significantly decrease the risk of abnormal serum ALT activity at the end of follow‐up (203/283 (71.7%) versus 218/280 (77.9%); RR 0.92, 95% CI 0.85 to 1.01). Six trials (Clerici 1994; Kawata 1994; Tanaka 1996; Kiso 1997; Viola 1998; Fabbri 2000) showed that long‐term treatment (12 months or more) with UDCA tended to significantly decrease the risk of abnormal serum ALT activity at the end of follow‐up (148/217 (68.2%) versus 166/217 (76.5%); RR 0.89, 95% CI 0.80 to 1.00) (Comparison 03‐09). No significant difference was observed between short‐term treatment and long‐term treatment (P = 0.67).

Serum ALT activity at the end of treatment   Two trials (Puoti 1995; Scotto 1997) compared UDCA versus placebo or no intervention and three trials including 224 patients (Crosignani 1998; Belloni 1999) compared TUDCA versus placebo or no intervention. Bile acids did not significantly decrease serum ALT activity at the end of treatment (WMD ‐1.29 IU/L, 95% CI ‐3.16 to ‐0.59) (Comparison 03‐10). However, there was significant heterogeneity in this comparison (Chi‐square = 69.55, df = 4, P < 0.0001). In subgroup analyses, UDCA (WMD ‐27 IU/L, 95% CI ‐40 to ‐14) but not TUDCA (WMD ‐0,73 IU/L, 95% CI ‐2.63 to ‐1.16) significantly decreased serum ALT activity at the end of treatment. The subgroup analysis regarding UDCA had significant heterogeneity (Chi‐square = 15.73, df = 1, P = 0.0001).

Serum AST activity at the end of treatment   One trial (Scotto 1997) compared UDCA versus placebo and one trial (Crosignani 1998) compared TUDCA versus placebo. Bile acids significantly decreased serum AST activity at the end of treatment (WMD ‐25 IU/L, 95% CI ‐34 to ‐16) (Comparison 03‐11). Both UDCA (WMD ‐21 IU/L, 95% CI ‐32 to ‐9) and TUDCA (WMD ‐32 IU/L, 95% CI ‐46 to ‐18) significantly decreased serum AST activity at the end of treatment.

Serum GGT activity at the end of treatment   Three trials (Leri 1994; Puoti 1995; Scotto 1997) compared UDCA versus placebo or no intervention and two trials (Crosignani 1998; Belloni 1999) compared TUDCA versus placebo or no intervention. Bile acids significantly decreased serum GGT activity at the end of treatment (WMD ‐14 IU/L, 95% CI ‐17 to ‐11) (Comparison 03‐12). However, there was significant heterogeneity in this comparison (Chi‐square = 16.03, df = 4, P = 0.003). In subgroup analyses, both UDCA (WMD ‐13 IU/L, 95% CI ‐16 to ‐10) and TUDCA (WMD ‐31 IU/L, 95% CI ‐46 to ‐17) significantly decreased serum GGT activity at the end of treatment. The subgroup analysis regarding UDCA had significant heterogeneity (Chi‐square = 10.11, df = 2, P = 0.0064).

Liver histology   Three trials (Angelico 1995a; Boucher 1995; Fabbri 2000) compared the effects of UDCA plus interferon versus interferon on the portal and periportal inflammation score. At the end of treatment, UDCA significantly increased portal and periportal inflammation score (WMD 0.20, 95% CI 0.15 to 0.24) based on a fixed‐effect model, but not a random‐effects model (WMD 0.02, 95% CI ‐0.62 to 0.66) (Comparison 03‐13). There was significant heterogeneity in this comparison (Chi‐square = 9.58, df = 2, P = 0.0083). Two trials (Boucher 1995; Fabbri 2000) compared the effects of UDCA plus interferon versus interferon on the Knodell score (Knodell 1981). At the end of treatment, UDCA plus interferon significantly increased the Knodell score (WMD 0.20, 95% CI 0.08 to 0.31) (Comparison 03‐14).

Adverse events   No serious adverse events were reported in the included trials. UDCA did not significantly increase the proportion of adverse events (15/147 (10.2%) versus 11/107 (10.3%); RR 1.17, 95% CI 0.56 to 2.46) (Comparison 03‐15). Two trials (Viola 1998; Poupon 2000), which compared UDCA plus interferon versus interferon, reported adverse events (11/89 (12.4%) versus 11/90 (12.2%); RR 1.06, 95% CI 0.49 to 2.28). In Poupon 2000, five adverse events (diarrhea in all cases) were reported (three in the UCLA plus interferon group and two in the placebo plus interferon group). In Viola 1998, the adverse events were reported without providing the types. One trial (Takano 1994), which compared UDCA versus no intervention, reported adverse events (4/58 (6.9%) versus 0/17 (0%); RR 2.75, 95% CI 0.16 to 48.61), which included three patients with abdominal discomfort and one patient with icterus in the UDCA group.

Funnel plot asymmetry   A funnel plot assessing the trials effect estimates on the outcome of having serum HCV RNA at the end of treatment for patients with chronic hepatitis C revealed no significant funnel plot asymmetry (Intercept ‐0.276, 95% CI ‐1.702 to 1.150, P = 0.676).

Bile acids for chronic viral hepatitis among heart‐transplanted patients   There was one trial assessing the effects of UDCA on chronic viral hepatitis including B and C among 60 heart‐transplanted patients after one year treatment. No significant differences between UDCA and placebo were found in terms of all cause mortality (7/30 (23.2%) versus 3/30 (10%); RR 2.74, 95% CI 0.63 to 11.82), raised serum ALT (24/30 (80.0%) versus 25/30 (83.3%); RR 0.96, 95% CI 0.76 to 1.22), Knodell score improvement (at least one point) (6/30 (20.0%) versus 13/30 (43.3%); RR 0.46, 95% CI 0.20 to 1.05), and graft rejection (8/30 (26.7%) versus 3/30 (10.0%); RR 2.67, 95% CI 0.78 to 9.09) at the end of treatment.

Discussion

We identified only two trials evaluating UDCA for acute or chronic hepatitis B. However, we identified a large number of trials evaluating UDCA or TUDCA for chronic hepatitis C. We were unable to identify any randomised clinical trials evaluating bile acids for hepatitis A, acute hepatitis C, acute or chronic hepatitis D, or hepatitis E. None of the included trials could be considered of high methodological quality with low risk of bias, ie, having adequate generation of the allocation sequence, allocation concealment, double blinding, and follow‐up (Schulz 1995; Moher 1998; Kjaergard 2001). Accordingly, the findings of our review must be interpreted with caution. Furthermore, the observed effects of bile acids were small and affected only surrogate outcomes (Gluud 2007). We have no evidence that these effects can be translated into meaningful clinical outcomes for patients. However, our findings suggest that UDCA may normalise or decrease serum transaminase activities in patients with acute or chronic hepatitis B and chronic hepatitis C. One single trial observed beneficial effects of UDCA on HBV markers in patients with acute hepatitis B. Further, a subgroup analysis of the trials evaluating long‐term UDCA treatment of chronic hepatitis C suggested a beneficial effect on clearance of HCV RNA. However, we need large randomised clinical trials conducted with adequate methodology to confirm these observations. We were mostly unable to identify data on mortality, liver related morbidity (ie, hepatocellular carcinoma and end‐stage liver disease), quality of life, or cost‐effectiveness. Bile acid treatment appeared safe; no serious adverse events were reported. Most of the adverse events which caused withdrawal of patients, such as depression, weight loss, and severe flu‐like syndrome, probably arose from co‐intervention with interferon (Brok 2005).

Bile acids for hepatitis B   One trial in patients with acute hepatitis B reported that UDCA significantly reduced the risk of having positive HBsAg at the end of treatment (but not the end of follow‐up) and significantly decreased HBV DNA levels at the end of follow‐up (but not the end of treatment). This observation requires replication for several reasons. First, UDCA had no consistent effects on serum enzyme activities at the end of treatment or follow‐up. Second, generation of the allocation sequence and allocation concealment were considered inadequate in this trial. Third, although the trial used placebo, we cannot be sure that this resulted in adequate blinding. Since trials involving bile acids may be difficult to blind sufficiently (Gluud 2001b), it cannot be excluded that the trial may have been biased. Therefore, the observations of Galský1999 need confirmation before any therapeutic recommendation can be made.

One trial (Zhou 1995) compared UDCA versus no intervention for patients with chronic hepatitis B. In this trial, UDCA significantly decreased the risk of abnormal serum ALT and serum ALT activities at the end of treatment. However, the trial did not provide evidence for an effect of UDCA on serum viral markers or clinical or histological outcomes. Therefore, we also need more trials to investigate the effects of UDCA for chronic hepatitis B.

Bile acids for chronic hepatitis C   Overall, bile acids did not significantly affect the risk of being HCV RNA positive at the end of treatment or at the end of follow‐up. In a sensitivity analysis considering treatment duration, bile acid treatment for 12 months or more was associated with a significant decrease in the risk of having HCV RNA at the end of follow‐up, but not at the end of treatment. Before considering therapeutic actions based on this finding, one should consider the following. First, why should bile acids, which are quickly metabolised (Fuchs 1999), affect HCV RNA levels at the end of follow‐up, but not at the end of treatment? Second, subgroup results, particularly when post hoc in nature, should be evaluated cautiously (Hahn 2000). Third, due to the low methodological quality of the trials, significant overestimation of intervention effects cannot be excluded (Schulz 1995; Moher 1998; Kjaergard 2001). Our observation should lead to proper assessment of bile acids in large scale randomised trials using adequate methods to control bias. The quest for such trials is further supported by the significant effects of bile acids on transaminase activities, whether analysed at the end of treatment or at the end of follow‐up.

Although we observed significant effects of UDCA on biochemical variables such as ALT, AST, and GGT, we could not find evidence to prove that the improvement of these variables could be translated into meaningful improvement of clinical outcomes such as mortality, incidence of cirrhosis, and hepatocellular carcinoma. In clinical practice, some physicians continue to base therapeutic decisions on surrogate measures such as the biochemical variables mentioned above (Gluud 2007). The results of our present review are comparable to two Cochrane systematic reviews on the effect of bile acids for primary biliary cirrhosis and primary sclerosing cholangitis, which reported significant improvement of biochemical variables, but not on 'hard' outcomes, such as mortality and liver transplantation (Gluud 2001b; Chen 2003). A new trial to investigate the effects of UDCA on meaningful clinical outcomes seems to be needed. Quality of life is another very important aspect that needs to be studied in patients with viral hepatitis. None of the included trials in our review assessed quality of life specifically.

Chronic hepatitis C infection can be characterised by histological changes, which run from relatively mild hepatic inflammatory activity and a low degree of fibrosis to full blown cirrhosis with liver failure. Hepatic lesions may be accompanied by bile duct damage, intraportal lymphoid aggregates, steatosis, or a combination of these manifestations (Mihm 1997). The pathological changes of the bile duct system caused by viral hepatitis contribute the formation of cholangitis and accumulation of hydrophobic bile acids in the liver tissue (Rodrigues 2000). Therefore, one possible explanation for the liver biochemistry improvement obtained by administering UDCA would be that it replaces the more hydrophobic human bile acids, which cause liver cell damage (Batta 1989; Chretien 1989). The protection of UDCA may also result from stimulation of hepato‐biliary secretion (Paumgartner 2002). Furthermore, recent data from basic research demonstrate that bile acids, including UDCA, appear to affect both 'death receptors' (Caspase 8/10) and 'cell survival cascades' (UDCA‐epidermal growth factor‐mitogen activated protein kinase) (Guicciardi 2002; Qiao 2002). The beneficial effects of UDCA on liver biochemistry variables suggest that in patients with viral hepatitis (the present review), primary biliary cirrhosis (Gluud 2001b), and primary sclerosing cholangitis (Chen 2003), UDCA's stimulus of the 'cell survival cascades' seems to overpower UDCA's stimulus of the cell 'death receptors'. However, beneficially affecting cell death of, for example hepatocytes, may not be important enough to arrest or slow down the progression of these liver diseases. From the meta‐analyses in our review, we observed that UDCA did have significant effects on improving the biochemical variables such as ALT, AST and GGT, which may result from the mechanisms of UDCA mentioned above. However, in this present review and another review on UDCA for primary biliary cirrhosis (Gluud 2001b), we found that bile acids may worsen liver histology. More evidence regarding this finding is needed.

As indicated, the reported methodological quality of the reviewed trials was low and in some trials we had difficulties extracting the relevant data. In order to get detailed data on the trials, 17 letters were sent to the first authors. Only one author replied and this author could not supply the information because the original data were lost. In the future, trials ought to report more fully on clinically relevant outcomes and follow recommended guidelines for the reporting of trials (CONSORT ‐ Consolidated Standards of Reporting Trials: www.consort‐statement.org).

We did not observe any therapeutic advantage of using TUDCA instead of UDCA. We therefore advocate that UDCA should be the bile acid being evaluated for patients with viral hepatitis, as this bile acid has been most widely used. Accordingly, we know the adverse event profile of this bile acid best. If it turns out to be possible to demonstrate beneficial clinical effects with UDCA, it may then become worthwhile considering 'head‐to‐head' comparisons of UDCA versus TUDCA in randomised clinical trials.

UDCA is still being used for primary biliary cirrhosis, although the evidence for its clinical effects have been questioned in a meta‐analysis (Goulis 1999) and in a Cochrane review (Gluud 2001b). In primary biliary cirrhosis, UDCA compared to placebo/no intervention leads to a decrease in serum bilirubin of about 10 µmol/L and a decrease of serum ALT of about 48 IU/L (Gluud 2001b). These decreases are similar to those observed in the present review in patients with chronic hepatitis B or C considering the dosage administered and the duration of treatment. Bile acids do not seem to differ regarding the lack of significant clinical effects in both primary biliary cirrhosis and viral hepatitis. It is, therefore, a question why we as clinicians are not more consistent as UDCA is often used for primary biliary cirrhosis and it is very seldom that UDCA (or other bile acids) are used for viral hepatitis.

This review demonstrates the necessity to search for trials in databases that are not normally included in Cochrane Reviews and other reviews. We identified five Chinese trials by searching the Chinese Biomedical Database. Although the methodological quality of Chinese clinical trials is generally low (Liu 2002), the rising activity in performing randomised clinical trials in China must be considered in the West.

In the above meta‐analyses we have disregarded the risks of random errors in meta‐analysis. Evidence shows that much errors may be substantial (Wetterslev 2007). If these risks were taken properly into consideration, several of our significant findings are likely to come out insignificant (Wetterslev 2007).

Authors' conclusions

Implications for practice.

There is not significant evidence to support or refute beneficial effects of bile acids for viral hepatitis. UDCA or TUDCA may improve liver transaminase activities, but there is no compelling evidence showing that these bile acids beneficially affect viral markers, mortality, cirrhosis development, need for liver transplantation, or liver histology in patients with acute or chronic hepatitis B and chronic hepatitis C. We have no knowledge about the potential beneficial and harmful effects of bile acids for hepatitis A, acute hepatitis C, hepatitis D, or hepatitis E.

Implications for research.

Acute hepatitis B   Due to the potential beneficial effect of UDCA on viral markers in acute hepatitis B, it seems worthwhile to test this in randomised placebo controlled clinical trials with sufficient power and long‐term follow‐up after the end of treatment. If such trials should show positive results, further trials addressing clinical effects seem warranted.

Chronic hepatitis B   Due to the potential beneficial effect of UDCA on viral markers in acute hepatitis B and on liver biochemistry in chronic hepatitis B, it seems worthwhile to test if similar effects could be observed in chronic hepatitis B patients included in randomised placebo controlled clinical trials with sufficient power and long‐term follow‐up after the end of treatment. Co‐intervention in both arms of the trial could be used. If such trials should show positive results on viral markers, further trials addressing clinical effects seem warranted.

Chronic hepatitis C   Due to the potential beneficial effect of bile acids on viral markers in chronic hepatitis C with co‐intervention of interferon, it seems worthwhile to test this potential effect in randomised placebo controlled clinical trials with sufficient power and long‐term follow‐up after the end of treatment. Co‐intervention in both arms of the trial could be pegylated interferon plus ribavirin. If such trials should show positive results, further trials addressing clinical effects seem warranted.

Other types of viral hepatitis   We suggest that trials on bile acids for other forms of viral hepatitis should await the results of trials demonstrating clear virological and/or clinical effects of bile acids for patients with hepatitis B and/or C.

General aspects   We do not have any data examining the potential advantage of using TUDCA instead of UDCA. However, before such 'head‐to‐head' comparative trials are considered, the beneficial clinical effects of UDCA ought to be established.

Researchers wishing to examine the effects of bile acids for viral hepatitis ought to use adequate trial methodology and follow the guidelines for reporting of trials (CONSORT ‐ Consolidated Standards of Reporting Trials: www.consort‐statement.org).

Detrimental effect of bile acids on liver histology cannot be excluded. Therefore, an independent Data Monitoring and Safety Committee should effectively monitor such trials, including the results of liver histology.

What's new

Date	Event	Description
17 October 2008	Amended	Converted to new review format.

Data base	Search strategy	Date of search
The Cochrane Hepato‐Biliary Group Controlled Trials Register	#1: 'hepatitis' and 'chenodeoxycholic'   #2: 'hepatitis' and 'cholic'   #3: 'hepatitis' and 'deoxycholic'   #4: 'hepatitis' and 'glycochenodeoxycholic'   #5: 'hepatitis' and 'glycocholic'   #6: 'hepatitis' and 'glycodeoxycholic'   #7: 'hepatitis' and 'glycolithocholic'   #8: 'hepatitis' and 'hyodeoxycholic'   #9: 'hepatitis' and 'lithocholic'   #10: 'hepatitis' and 'taurochenodeoxycholic'   #11: 'hepatitis' and 'taurocholic'   #12: 'hepatitis' and 'taurodehydrocholic'   #13: 'hepatitis' and 'taurodeoxycholic'   #14: 'hepatitis' and 'tauroglycocholic'   #15: 'hepatitis' and 'taurolithocholic'   #16: 'hepatitis' and 'tauroselcholic'   #17: 'hepatitis' and 'tauroursocholic'   #18: 'hepatitis' and 'tauroursodeoxycholic'   #19: 'hepatitis' and 'ursocholic'   #20: 'hepatitis' and 'ursodeoxycholic'		July 2007.
The Cochrane Controlled Trials Register	#1: HEPATITIS:ME   #2: HEPATITIS   #3: BILE ACID :ME   #4: CHENODEOXYCHOLIC or CHOLIC or DEOXYCHOLIC or GLYCOCHENODEOXYCHOLIC or GLYCOCHOLIC or GLYCODEOXYCHOLIC or GLYCOLITHOCHOLIC or HYODEOXYCHOLIC or LITHOCHOLIC or TAUROCHENODEOXYCHOLIC or TAUROCHOLIC or TAURODEHYDROCHOLIC or TAURODEOXYCHOLIC or TAUROGLYCOCHOLIC or TAUROLITHOCHOLIC or TAUROSELCHOLIC or TAUROURSOCHOLIC or TAUROURSODEOXYCHOLIC or URSOCHOLIC or URSODEOXYCHOLIC   #5: #1 or #2   #6: #3 or #4   #7: #5 and #6		July 2007.
MEDLINE	#1: HEPATITIS :ME   #2: HEPATITIS   #3: BILE ACID :ME   #4: CHENODEOXYCHOLIC or CHOLIC or DEOXYCHOLIC or GLYCOCHENODEOXYCHOLIC or GLYCOCHOLIC or GLYCODEOXYCHOLIC or GLYCOLITHOCHOLIC or HYODEOXYCHOLIC or LITHOCHOLIC or TAUROCHENODEOXYCHOLIC or TAUROCHOLIC or TAURODEHYDROCHOLIC or TAURODEOXYCHOLIC or TAUROGLYCOCHOLIC or TAUROLITHOCHOLIC or TAUROSELCHOLIC or TAUROURSOCHOLIC or TAUROURSODEOXYCHOLIC or URSOCHOLIC or URSODEOXYCHOLIC   #5: #1 or #2   #6: #3 or #4   #7: #5 and #6   #8: RANDOMIZED‐CONTROLLED‐TRIAL :ME   #9: RANDOM   #10: #8 or #9   #11: #7 and #10		July 2007.
EMBASE	#1: HEPATITIS :ME   #2: HEPATITIS   #3: BILE ACID :ME   #4: CHENODEOXYCHOLIC or CHOLIC or DEOXYCHOLIC or GLYCOCHENODEOXYCHOLIC or GLYCOCHOLIC or GLYCODEOXYCHOLIC or GLYCOLITHOCHOLIC or HYODEOXYCHOLIC or LITHOCHOLIC or TAUROCHENODEOXYCHOLIC or TAUROCHOLIC or TAURODEHYDROCHOLIC or TAURODEOXYCHOLIC or TAUROGLYCOCHOLIC or TAUROLITHOCHOLIC or TAUROSELCHOLIC or TAUROURSOCHOLIC or TAUROURSODEOXYCHOLIC or URSOCHOLIC or URSODEOXYCHOLIC   #5: #1 or #2   #6: #3 or #4   #7: #5 and #6   #8: RANDOMIZED‐CONTROLLED‐TRIAL :ME   #9: RANDOM   #10: #8 or #9   #11: #7 and #10		July 2007.
The Chinese Biomedical Database	#1: HEPATITIS :ME   #2: HEPATITIS   #3: BILE ACID :ME   #4: CHENODEOXYCHOLIC or CHOLIC or DEOXYCHOLIC or GLYCOCHENODEOXYCHOLIC or GLYCOCHOLIC or GLYCODEOXYCHOLIC or GLYCOLITHOCHOLIC or HYODEOXYCHOLIC or LITHOCHOLIC or TAUROCHENODEOXYCHOLIC or TAUROCHOLIC or TAURODEHYDROCHOLIC or TAURODEOXYCHOLIC or TAUROGLYCOCHOLIC or TAUROLITHOCHOLIC or TAUROSELCHOLIC or TAUROURSOCHOLIC or TAUROURSODEOXYCHOLIC or URSOCHOLIC or URSODEOXYCHOLIC   #5: #1 or #2   #6: #3 or #4   #7: #5 and #6   #8: RANDOMIZED‐CONTROLLED‐TRIAL :ME   #9: RANDOM   #10: #8 or #9   #11: #7 and #10		July 2007.
Science Citation Index Expanded	#1: HEPATITIS :ME   #2: HEPATITIS   #3: BILE ACID :ME   #4: CHENODEOXYCHOLIC or CHOLIC or DEOXYCHOLIC or GLYCOCHENODEOXYCHOLIC or GLYCOCHOLIC or GLYCODEOXYCHOLIC or GLYCOLITHOCHOLIC or HYODEOXYCHOLIC or LITHOCHOLIC or TAUROCHENODEOXYCHOLIC or TAUROCHOLIC or TAURODEHYDROCHOLIC or TAURODEOXYCHOLIC or TAUROGLYCOCHOLIC or TAUROLITHOCHOLIC or TAUROSELCHOLIC or TAUROURSOCHOLIC or TAUROURSODEOXYCHOLIC or URSOCHOLIC or URSODEOXYCHOLIC   #5: #1 or #2   #6: #3 or #4   #7: #5 and #6   #8: RANDOMIZED‐CONTROLLED‐TRIAL :ME   #9: RANDOM   #10: #8 or #9   #11: #7 and #10		July 2007.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Risk of having positive serum HBsAg at the end of treatment	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.40 [0.17, 0.92]
1.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.40 [0.17, 0.92]
2 Risk of having positive HBsAg at the end of follow‐up	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.23 [0.05, 1.00]
2.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.23 [0.05, 1.00]
3 Serum DNA (pg/ml) level at the end of treatment	1	59	Mean Difference (IV, Fixed, 95% CI)	‐574.05 [‐1148.71, 0.61]
3.1 UDCA versus placebo	1	59	Mean Difference (IV, Fixed, 95% CI)	‐574.05 [‐1148.71, 0.61]
4 Serum DNA (pg/ml) level at the end of follow‐up	1	59	Mean Difference (IV, Fixed, 95% CI)	‐954.73 [‐1813.71, ‐95.75]
4.1 UDCA versus placebo	1	59	Mean Difference (IV, Fixed, 95% CI)	‐954.73 [‐1813.71, ‐95.75]
5 Risk of abnormal serum ALT at the end of treatment	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.35 [0.12, 1.02]
5.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.35 [0.12, 1.02]
6 Risk of abnormal serum ALT at the end of follow‐up	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.26 [0.06, 1.21]
6.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.26 [0.06, 1.21]
7 Risk of abnormal serum GGT at the end of treatment	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.32 [0.11, 0.90]
7.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.32 [0.11, 0.90]
8 Risk of abnormal serum GGT at the end of follow‐up	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.26 [0.06, 1.21]
8.1 UDCA versus placebo	1	61	Risk Ratio (M‐H, Fixed, 95% CI)	0.26 [0.06, 1.21]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Risk of abnormal serum ALT at the end of treatment	1	112	Risk Ratio (M‐H, Fixed, 95% CI)	0.65 [0.45, 0.94]
1.1 UDCA versus no intervention	1	112	Risk Ratio (M‐H, Fixed, 95% CI)	0.65 [0.45, 0.94]
2 Serum ALT (IU/L) at the end of treatment	1	112	Mean Difference (IV, Fixed, 95% CI)	‐14.94 [‐21.20, ‐8.68]
2.1 UDCA versus no intervention	1	112	Mean Difference (IV, Fixed, 95% CI)	‐14.94 [‐21.20, ‐8.68]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Risk of having serum HCV RNA at the end of treatment	13	837	Risk Ratio (M‐H, Fixed, 95% CI)	0.99 [0.91, 1.07]
1.1 UDCA versus placebo or no intervention	2	106	Risk Ratio (M‐H, Fixed, 95% CI)	1.01 [0.84, 1.21]
1.2 UDCA plus interferon versus interferon	11	731	Risk Ratio (M‐H, Fixed, 95% CI)	0.98 [0.90, 1.08]
2 Sensitivity analyses: Risk of having serum HCV RNA at the end of treatment ‐ Duration of treatment	13	837	Risk Ratio (M‐H, Fixed, 95% CI)	0.99 [0.91, 1.07]
2.1 Short treatment duration (less than 12 months)	8	444	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.83, 1.04]
2.2 Long treatment duration (12 months or more)	5	393	Risk Ratio (M‐H, Fixed, 95% CI)	1.05 [0.92, 1.19]
3 Risk of having serum HCV RNA at the end of follow‐up	10	676	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.87, 1.00]
3.1 UDCA plus interferon versus interferon	10	676	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.87, 1.00]
4 Sensitivity analyses: Risk of having serum HCV RNA at the end of follow‐up ‐ Duration of treatment	10	676	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.87, 1.00]
4.1 Short treatment duration (less than 12 months)	5	283	Risk Ratio (M‐H, Fixed, 95% CI)	0.99 [0.87, 1.12]
4.2 Long treatment duration (12 months or more)	5	393	Risk Ratio (M‐H, Fixed, 95% CI)	0.90 [0.83, 0.99]
5 Risk of cirrhosis at the end of treatment	1	103	Risk Ratio (M‐H, Fixed, 95% CI)	1.79 [0.93, 3.47]
5.1 UDCA plus interferon versus interferon	1	103	Risk Ratio (M‐H, Fixed, 95% CI)	1.79 [0.93, 3.47]
6 Risk of abnormal serum ALT at the end of treatment	21	1582	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.77, 0.90]
6.1 UDCA versus placebo or no intervention	3	239	Risk Ratio (M‐H, Fixed, 95% CI)	0.78 [0.69, 0.88]
6.2 UDCA plus interferon versus interferon	15	1007	Risk Ratio (M‐H, Fixed, 95% CI)	0.81 [0.73, 0.91]
6.3 TUDCA plus interferon versus interferon	2	166	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.70, 1.24]
6.4 UDCA plus glycyrrhizin versus glycyrrhizin	1	170	Risk Ratio (M‐H, Fixed, 95% CI)	0.96 [0.75, 1.24]
7 Sensitivity analyses: Risk of abnormal serum ALT at the end of treatment ‐ Duration of treatment	21	1582	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.77, 0.90]
7.1 Short treatment duration (less than 12 months)	16	1189	Risk Ratio (M‐H, Fixed, 95% CI)	0.84 [0.76, 0.92]
7.2 Long treatment duration (12 months or more)	5	393	Risk Ratio (M‐H, Fixed, 95% CI)	0.83 [0.69, 0.99]
8 Risk of abnormal serum ALT at the end of follow‐up	13	929	Risk Ratio (M‐H, Fixed, 95% CI)	0.91 [0.85, 0.97]
8.1 UDCA plus interferon versus interferon	12	869	Risk Ratio (M‐H, Fixed, 95% CI)	0.90 [0.84, 0.97]
8.2 TUDCA plus interferon versus interferon	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	1.0 [0.68, 1.47]
9 Sensitivity analyses: Risk of abnormal serum ALT at the end of follow‐up ‐ Duration of treatment	13	929	Risk Ratio (M‐H, Fixed, 95% CI)	0.91 [0.85, 0.97]
9.1 Short treatment duration (less than 12 months)	7	495	Risk Ratio (M‐H, Fixed, 95% CI)	0.92 [0.84, 1.00]
9.2 Long treatment duration (12 months or more)	6	434	Risk Ratio (M‐H, Fixed, 95% CI)	0.89 [0.80, 1.00]
10 Serum ALT (IU/L) at the end of treatment	5	386	Mean Difference (IV, Fixed, 95% CI)	‐1.29 [‐3.16, 0.59]
10.1 UDCA versus placebo or no intervention	2	162	Mean Difference (IV, Fixed, 95% CI)	‐26.78 [‐39.65, ‐13.92]
10.2 TUDCA versus placebo or no intervention	3	224	Mean Difference (IV, Fixed, 95% CI)	‐0.73 [‐2.63, 1.16]
11 Serum AST (IU/L) at the end of treatment	2	146	Mean Difference (IV, Fixed, 95% CI)	‐25.16 [‐34.02, ‐16.30]
11.1 UDCA versus placebo or no intervention	1	61	Mean Difference (IV, Fixed, 95% CI)	‐20.70 [‐32.08, ‐9.32]
11.2 TUDCA versus placebo or no intervention	1	85	Mean Difference (IV, Fixed, 95% CI)	‐32.0 [‐46.10, ‐17.90]
12 Serum GGT (IU/L) at the end of treatment	5	302	Mean Difference (IV, Fixed, 95% CI)	‐14.09 [‐17.34, ‐10.84]
12.1 UDCA versus placebo or no intervention	3	184	Mean Difference (IV, Fixed, 95% CI)	‐13.14 [‐16.48, ‐9.80]
12.2 TUDCA versus placebo or no intervention	2	118	Mean Difference (IV, Fixed, 95% CI)	‐31.23 [‐45.45, ‐17.01]
13 Portal and periportal inflammation scores at the end of treatment	3	167	Mean Difference (IV, Fixed, 95% CI)	0.20 [0.15, 0.24]
13.1 UDCA plus interferon versus interferon	3	167	Mean Difference (IV, Fixed, 95% CI)	0.20 [0.15, 0.24]
14 Knodell score at the end of treatment	2	133	Mean Difference (IV, Fixed, 95% CI)	0.20 [0.08, 0.31]
14.1 UDCA plus interferon versus interferon	2	133	Mean Difference (IV, Fixed, 95% CI)	0.20 [0.08, 0.31]
15 Adverse events caused by bile acid	3	254	Risk Ratio (M‐H, Fixed, 95% CI)	1.17 [0.56, 2.46]
15.1 UDCA plus interferon versus interferon	2	179	Risk Ratio (M‐H, Fixed, 95% CI)	1.06 [0.49, 2.28]
15.2 UDCA versus no intervention	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	2.75 [0.16, 48.61]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 All cause mortality at the end of treatment	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	2.33 [0.67, 8.18]
2 Risk of abnormal serum ALT at the end of treatment	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	0.96 [0.76, 1.22]
3 Improvement of total Knodell score at the end of treatment	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	0.46 [0.20, 1.05]
4 Number of graft rejection at the end of treatment	1	60	Risk Ratio (M‐H, Fixed, 95% CI)	2.67 [0.78, 9.09]

Methods	Generation of allocation sequence: unclear (no description).   Allocation concealment: unclear (no description).   Double blinding: adequate (identical placebo).   Follow up: adequate (   UDCA plus IFN group (one patient);   placebo plus IFN group (one patient).   Sample size calculation: no information.   Intention‐to‐treat analysis: not performed.
Participants	Country: United States.   Type of hepatitis: chronic hepatitis C.   INCLUSION CRITERIA   ‐ outpatients;   ‐ persistent serum ALT elevation (at least three times the upper normal limit) on two or more occasions during six months before enrolment;   ‐ serum HCV RNA positive;   ‐ negative for HBsAg and HIV;   ‐ no history of exposure to hepatotoxins, alcohol intake greater than 20 g/day, or illicit drug use.   EXCLUSION CRITERIA   ‐ other causes of chronic liver disease;   ‐ decompensated liver disease, severe concomitant medical or psychiatric illness, malignancy, untreated thyroid disease, or renal insufficiency;   ‐ previous treatment with any type of IFN, UDCA, or immunosuppressive or antiviral agents within six months of entry into the study.   PARTICIPANTS   ‐ UDCA plus IFN group (n = 16):   Mean age (years +/‐ SD)   42.5 +/‐ 1.8.   Ratio of sex (M/F)   10/6.   ‐ Placebo plus IFN group (n = 15):   Mean age (years +/‐ SD)   46.2 +/‐ 3.1.   Ratio of sex (M/F)   11/4.   Proportion of patients with cholestasis: no information.   Proportion of patients with cirrhosis: no information.
Interventions	UDCA plus IFN group:   UDCA   ‐ Dose: 13‐15 mg/kg/day;   ‐ Route: orally;   ‐ Duration: six months.   Alpha IFN 2b   ‐ Dose: three million units three times a week;   ‐ Route: subcutaneously;   ‐ Duration: six months. Placebo plus IFN group:   Identical placebo.   Alpha IFN‐2b   ‐ Dose: three million units three times a week;   ‐ Route: subcutaneously;   ‐ Duration: six months.
Outcomes	‐ Normalisation of serum ALT at the end of treatment and follow‐up.   ‐ Decrease of at least three points in the Knodell score at the end of treatment.   ‐ Absence of HCV RNA in serum at the end of treatment and follow‐up.
Notes	Follow‐up time: six months after the end of treatment.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	Generation of allocation sequence: adequate   (computer‐generated randomisation list).   Allocation concealment: adequate (sealed envelopes).   Double blinding: inadequate (not blinding).   Follow up: adequate (one patient in each group).   Sample size calculation: yes.   Intention‐to‐treat analysis: yes.
Participants	Country: Italy.   Type of hepatitis: chronic hepatitis C.   INCLUSION CRITERIA   ‐ age between 18 and 65 years;   ‐ persistent serum ALT elevation during the six months before the study, with value three times above the upper normal limit for at least three months;   ‐ serum anti‐HCV positive;   ‐ histological evidence of chronic active or persistent hepatitis within three months of entry, with or without cirrhosis.   EXCLUSION CRITERIA   ‐ grade B and C cirrhosis, according to Child‐Pugh classification;   ‐ presence of HBsAg in serum;   ‐ recent use of corticosteroids or previous IFN medications;   ‐ recent history of alcohol or drug abuse;   ‐ presence of autoimmune, genetic, or other types of liver diseases;   ‐ platelet count below 100,000/µl and or leukocyte count below 3,000/µl;   ‐ malignancies or renal, hematological, cardiopulmonary, neurological, and gastrointestinal diseases;   ‐ marked abnormalities in liver function tests (serum bilirubin > 70 µmol/l and/or serum albumin < 30 g/l and/or prothrombin time < 50%);   ‐ serum positivity for anti‐HIV antibody.   PARTICIPANTS   ‐ UDCA plus IFN group (n = 20):   Mean age (years +/‐ SD)   49 +/‐ 14.   Ratio of sex (M/F) 18/2.   Proportion of patients with cholestasis: no information.   Proportion of patients with cirrhosis: no information.   ‐ IFN group (n = 20):   Mean age (years +/‐ SD)   44 +/‐ 15.   Ratio of sex (M/F)   16/4.   Proportion of patients with cholestasis: no information.   Proportion of patients with cirrhosis: no information.
Interventions	UDCA plus IFN group:   UDCA   ‐ Dose: 10 mg/kg/day;   ‐ Route: orally;   ‐ Timing: two times a day;   ‐ Duration: nine months and 15 days.   Alpha IFN 2a   ‐ Dose: 3 million units three times a week;   ‐ Route: subcutaneously;   ‐ Duration: six months. IFN group:   Alpha IFN‐2a   ‐ Dose: three million units three times a week;   ‐ Route: subcutaneously;   ‐ Duration: six months.
Outcomes	‐ Normalisation of serum ALT at the end of treatment and follow‐up.   ‐ Absence of HCV RNA in serum at the end of treatment and follow‐up.   ‐ Histological changes six months after cessation of IFN therapy.
Notes	Follow‐up time: 18 months after stopping alpha IFN treatment.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Study	Reason for exclusion
Angelico 1995b	This study is a randomised cross‐over clinical trial. Forty anti‐HCV positive patients were randomised to receive IFN alpha, from day 0 to 180, plus TUDCA, from day 61 to 240; or TUDCA, from day 0 to 180, plus IFN alpha from day 61 to 240. The two interventions did not differ significantly. The trial does not fulfill our inclusion criteria.
Attilli 1994	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Thirty‐six patients with chronic active hepatitis were included. Patients were randomly allocated to receive 600 mg UDCA per day or placebo. Clinical and biochemical follow‐up was performed at three‐month intervals. A percutaneous liver biopsy was performed before and after one year of treatment. Multifactorial covariance analysis showed that the reductions in ALT, AST, and GGT were significantly higher in UDCA group than in the placebo group. Biochemical remission was not observed in either group. No differences were found in the two groups before or after treatment in histological activity index.
Buzzelli 1991	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. A total of 40 patients with chronic active hepatitis were randomised to oral UDCA 300 mg two times a day or S‐adenosyl‐methionine (200 mg two times a day) for six months. Treatment with UDCA produced a statistically significant reduction in ALT, AST, and GGT compared with SAMe.
Buzzelli 1992	This study is a randomised clinical trial, but the type of viral hepatitis can not be identified. A total of 36 patients with chronic hepatitis B and/or C were included. Eighteen were randomised to 300 mg of UDCA‐hemisuccinate orally twice a day for six months and 18 randomised to 200 mg of S‐adenosyl‐methionine twice a day for six months. Treatment with UDCA produced a statistically significant reduction in ALT, AST and GGT compared with SAMe.
Crosignani 1991	This study is not a randomised clinical trial, but a case series. The effects of UDCA on liver function tests and on bile acid metabolism were investigated in 18 patients with chronic active hepatitis. Three different doses of UDCA ‐ 250 mg, 500 mg, and 750 mg ‐ were administered daily to each patient for two months. A significant decrease in serum aminotransferase occurred with the lowest dose of UDCA, which corresponded to four mg/kg body weight/day, and no further significant decrease in serum aminotransferases with the higher doses was seen.
Crosignani 1998(a)	This study is a randomised clinical trial, but the type of viral hepatitis can not be identified. One hundred and fifty‐five patients with chronic active hepatitis were randomly assigned to receive TUDCA at the daily doses of 250, 500, and 1000 mg, or no treatment for six months. Serum aminotransferase and GGT activities decreased with each dose of TUDCA compared with controls (P < 0.001). The 1000 mg dose was followed by more marked improvement compared with the 250 mg dose (P < 0.05). An improvement of ALT with time (P < 0.05) was found only with the two higher doses.
Del Vecchio 1982	This study is a randomised clinical trial, but the type of viral hepatitis can not be identified. Forty‐four in‐patients with low activity chronic hepatitis were included in this study. Patients were randomly allocated to receive in a double‐blind way UDCA (10 mg/kg/day) or placebo in divided doses for three months. In both groups there was a significant improvement in all biochemical variables but without a statistically significant difference between treatments.
Italian 1990	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Sixty‐three chronic active hepatitis patients were enrolled in three participating centres: 34 received UDCA and 29 placebo. Serum AST and GGT activities were significantly reduced (P < 0.05 and P < 0.01, respectively) during the treatment with UDCA but not with placebo. Liver histology remained substantially unchanged in terms of periportal necrosis, intralobular degeneration, portal inflammation, and fibrosis in patients treated with UDCA or placebo.
Kadayifci 1997	This study is a case‐report. A 38‐year‐old man had pruritus and jaundice of eight weeks, initially compatible with acute hepatitis B. The patient took 10 mg/kg bodyweight/day UDCA, and the symptoms began to improve after two weeks. Bilirubin levels also gradually reduced. After 14 weeks of UDCA therapy, bilirubin levels and all other laboratory parameters returned to normal, hepatitis B surface antigen disappeared, and hepatitis B antibody became positive.
Lu 1995	This study is not a randomised clinical trial, but a case‐control study. It was conducted to evaluate the efficacy of UDCA in the treatment of Chinese patients with chronic hepatitis C. Patients who failed to have sustained responses to IFN therapy, refused to take IFN, or were unsuitable for IFN treatment, were enrolled. Fifteen patients received UDCA 600 mg orally per day for six months. Another fifteen patients were chosen as the control group. After the treatment period, the mean serum ALT activities in both groups were not significantly different and mean serum ALT activities in the UDCA‐treated group did not decrease after the treatment.
Pinto 1992	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Forty patients (15 M, 25 F) with biopsy proven chronic liver disease were randomly allocated to two treatment groups. Twenty patients received UDCA 600 mg/day and 20 patients received placebo for six months. Serum ALT activities were significantly reduced in the treated group when compared with placebo group at the end of treatment.
Podda 1989	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Forty‐eight patients (30 with PBC, six with PSC and 12 with chronic hepatitis) were included in this trial. UDCA was administered at dosages of 250, 500, and 750 mg/day for two months. Highly significant decreases in serum aminotransferase activities were observed in all groups, but there is no significant difference for the decrease in serum aminotransferase activities between the 500 and the 750 mg/day doses.
Podda 1990	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. The trial compared the effects of UDCA, taurine, or a combination of the two on indices of liver injury in 24 patients with chronic active hepatitis. They were assigned at random to two of the four following treatments: UDCA (600 mg/day), taurine (1.5 g/day), UDCA (600 mg/day) plus taurine (1.5 g/day) or placebo, given in two successive cycles of two months each. UDCA and the combination of UDCA and taurine significantly decreased serum AST, ALT, and GGT at the end of treatment when compared with taurine alone.
Podda 1995	This study is a randomised clinical trial, but the type of viral hepatitis can not be identified. In seven Italian centres, 155 patients with histological diagnosis of chronic active hepatitis were enrolled. They were randomly assigned to receive a six‐months course of TUDCA at the daily doses of 250 mg, 500 mg, 1000 mg, or no treatment. After six months of treatment, AST, ALT, and GGT decreased in patients administered TUDCA compared to patients receiving no treatment (P < 0.001).
Portincasa 1993	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. A total of 53 patients with histologic evidence of chronic active hepatitis were enrolled in the study. TUDCA 500 mg/day divided into two doses with meals was given to 27 patients; 26 patients served as controls. Follow‐ups were performed for one month, two months, and the end of the study period. TUDCA significantly lowered AST (‐44%), ALT (‐49%), and GGT(‐38%). Throughout the study, serum levels of alkaline phosphatases and serum bilirubin concentration remained within normal range in all patients.
Rolandi 1991	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Twenty‐six patients with serum ALT values at least twice the upper normal limit in two of three pre‐treatment tests received UDCA 450 mg/day or a placebo for twelve weeks. In all UDCA‐treated patients, serum AST, ALT, GGT and alkaline phosphatases fell significantly after four weeks of treatment when compared with patients in placebo group. Four weeks after suspension of therapy, there was no significant difference between UDCA group and placebo group in serum aminotransferase.
Song 1998	This study is a randomised clinical trial, but the type of viral hepatitis cannot be identified. Seventy‐nine patients with chronic active hepatitis were divided into two groups. Forty‐two patients in the treatment group received UDCA 300‐450 mg a day and thirty‐seven patients in control group received no treatment. The treatment group improved the serum GGT activities significantly when compared to the control group.
Zhu 1997	This study is not a randomised clinical trial, but a control series. Fifty‐three patients with chronic hepatitis C received UDCA 600 mg a day with the combination of ribaviram 900 mg a day and polyinosine polycytidylic acid four mg once in two days. Forty‐seven patients with chronic hepatitis C received ribaviram 900 mg a day and polyinosine polycytidylic acid four mg once in two days as control group. A significant improvement on biochemical response at the end of treatment in UDCA group was observed when compared with the control group.

PERMALINK

Bile acids for viral hepatitis

Wendong Chen

Jian Ping Liu

Christian Gluud

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Search methods for identification of studies

Data collection and analysis

Results

Description of studies

Risk of bias in included studies

Effects of interventions

Discussion

Authors' conclusions

Implications for practice.

Implications for research.

What's new

Notes

Acknowledgements

Appendices

Appendix 1. Search Strategies

Data and analyses

Comparison 1. Bile acids for acute hepatitis B.

1.1. Analysis.

1.2. Analysis.

1.3. Analysis.

1.4. Analysis.

1.5. Analysis.

1.6. Analysis.

1.7. Analysis.

1.8. Analysis.

Comparison 2. Bile acids for chronic hepatitis B.

2.1. Analysis.

2.2. Analysis.

Comparison 3. Bile acids for chronic hepatitis C.

3.1. Analysis.

3.2. Analysis.

3.3. Analysis.

3.4. Analysis.

3.5. Analysis.

3.6. Analysis.

3.7. Analysis.

3.8. Analysis.

3.9. Analysis.

3.10. Analysis.

3.11. Analysis.

3.12. Analysis.

3.13. Analysis.

3.14. Analysis.

3.15. Analysis.

Comparison 4. Bile acids for chronic viral hepatitis among heart transplanted patients.

4.1. Analysis.

4.2. Analysis.

4.3. Analysis.

4.4. Analysis.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Abdelmalek 1998.

Angelico 1995a.

Belloni 1999.

Bonnand 1996.

Boucher 1995.

Cadranel 2003.

Clerici 1994.