The effect of ursodeoxycholic acid on serum enzymes and liver histology in patients with chronic active hepatitis

Journal o f Hepatology 1994; 20:315-320 PrOtted 07 Denmark. All rights reserved Munksgaard. Copenhagen

Copyright © Journalof Hepatology 1994 Journal of Hepatology ISSN 0168-8278

The effect of ursodeoxycholic acid on serum enzymes and liver histology in patients with chronic active hepatitis A 12-month double-blind, placebo-controlled trial Adolpo Francesco Attili, Annabiulia Rusticali, Maria Varriale t, Laura Carli ~, Anna Maria Repice ~ and Francesco Callea-' Cattedra di Gastroenterologia, Universitt'z di L 'Aquila, ~Cattedra di Gastroenterologia, Universitgt di Roma "La Sapienza" and :Servizio di Anatomia Patologica, Ospedale Gaslini di Genova, Ball,

(Received 6 May 1991)

A reduction in serum enzymes has been already observed by administering ursodeoxycholic acid to patients with chronic active hepatitis. The aim of this study was to assess whether the liver histological activity of inflammation was modified by a 12-month treatment with ursodeoxycholic acid. Thirty-six patients with chronic active hepatitis, fulfilling the inclusion criteria, were admitted to the trial. Patients were randomly allocated to receive double blind either 600 mg/day of ursodeoxycholic acid (Group A: 18 patients) or placebo (Group B: 18 patients). Clinical and biochemical follow-up was performed at 3-month intervals. A percutaneous liver biopsy was performed before and after 1 year of treatment. Histological hepatitis activity was assessed using Knodell's numerical scoring system, while biliary damage was evaluated by an appropriate scoring system. Sixteen and 12 patients in Groups A and B, respectively, completed the clinical and biochemical follow-up. Although a reduction in serum enzymes was found in both groups, multifactorial covariance analysis showed that the reductions in alanine aminotransferase, aspartate aminotransferase and gamma glutamyl transpeptidase were significantly higher in Group A than in Group B. Biochemical remission was not observed in either group. Histological analysis showed a dichotomy between the results from the hepatitis and the biliary components of the disease process. No differences were found in the two groups before or after treatment in histological activity index, which measures the "hepatitic" component. Nor were there any significant differences in baseline values. Furthermore, no relationship was found between changes in biochemical parameters and the histological activity indices. In contrast, the "biliary" component showed significant improvement after ursodeoxycholic acid treatment for certain cholestatic parameters: ductular metaplasia, bile duct damage and phenotypic cytokeratin changes. These changes were significantly associated with changes in gamma glutamyl transpeptidase serum levels. This study has confirmed that ursodeoxycholic acid administration in patients with chronic active hepatitis reduces transaminases and gamma glutamyl transpeptidase. The morphological substrate of this biochemical improvement can be traced to the biliary component of the process. © Journal of Hepatology. K e y words: Cholestasis; Liver disease

Japanese investigators were first to report an improvement in liver function tests during ursodeoxycholic acid (UDCA) treatment in patients with chronic active hepatitis (1-5). Several years later, Leushner et al. (6) reported a decrease in serum transaminases in six patients with chronic active hepatitis and gallstones when they were treated with UDCA.

In a recent study we hypothesized that liver damage might be related to the hydrophobic-hydrophilic bile acid balance (7). In addition to conditions where the liver is perfused by highly detergent endogenous or exogenous bile acids (chenodeoxycholic or deoxycholic acids), a normally detergent bile acid pool could also become hepatotoxic for injured liver cells. Liver toxicity induced by bile

Correspondence to: A. F. Attili, MD, Cattedra di Gastroenterologia, Dipartimento di Medicina Interna, via S Sisto 20, 67100 L'Aquila, Italy Dedication: This paper is dedicated to Professor Gustav Paumgartner on the occasion of his 60th birthday.

316 salt detergency could be prevented by favouring tauroconjugation, reducing intestinal degradation of bile salts or possibly by administering low detergent bile acids. A significant reduction in serum transaminases was observed in chronic active hepatitis patients during treatment with taurine (8), while Podda and coworkers failed to observe similar results (9,10). No data are available on the effects of measures such as the administration of nonabsorbable antibiotics or cathartics which lower the intestinal degradation of bile acids and the formation of secondary, toxic, bile acids. Double-blind, placebo-controlled studies have reported a significant improvement in serum transaminases and gamma glutamyl transpeptidase (GGT) in patients with chronic active hepatitis taking UDCA (10,1 I). The aim of this study was to assess the effect of UDCA on serum enzymes and liver histology in patients with chronic active hepatitis.

Patients and Methods

Thirty-six patients, who fulfilled the inclusion criteria and were consecutively admitted to our gastrointestinal unit between June 1985 and January 1989, participated in the study. The study protocol was approved by the local Ethics Committee. Criteria for inclusion were the presence of histological features of chronic active hepatitis, with or without cirrhosis. Diagnosis was made in all cases during hospitalization, using percutaneous liver biopsy with a modified Menghini needle (Surecut 16G). The second criterion was the presence of elevated aminotransferase serum levels (more than 2.5 times above the upper limits of normal values), for 6 months or longer, prior to liver biopsy. None of the patients was excluded on the basis of exclusion criteria: under 12 years of age, presence of neoplasms, encephalopathy, ascites or renal insufficiency, anti-nuclear antibody, anti-mitochondrial antibody, liverkidney microsomal antibody or anti-smooth muscle antibody (actin) positivity. This double-blind study was carried out over a period of 12 months. The 36 patients admitted to the study were randomized to receive either UDCA 600 rag/day (300 mg b.i.d.) (Group A: 18 patients) or capsules containing a placebo (Group B: 18 patients), which were identical to those given to Group A. Randomization was performed using an SAS programme (SAS User's Guide, Version 5, SAS Institute Inc. Cary NC USA, 1983) which generates random numbers of permuted blocks within treatment. Ursodeoxycholic acid was kindly supplied by SanofiWinthrop. Clinical and biochemical follow-up examinations were performed at 3, 6, 9, and 12 months. Biochemical remission was said to have occurred if serum

A.F. ATTILI et al. aspartate (AST) and alanine aminotransferase (ALT) and gamma glutamyl transpeptidase were restored to normal values. Treatment compliance was cross-checked with a questionnaire and by counting the number of capsules remaining after each 3-month treatment period. No side effects were observed. A second percutaneous liver biopsy was performed after 12 months of treatment. Liver biopsies were independently and blindly assessed by a member of our team (C.F.). The histological analysis evaluated both hepatitis activity and biliary morphological changes. Hepatitis activity was assessed according to Knodell's scoring system (12), which is based on individual and cumulative considerations of four histological parameters: periportal___bridging necrosis, intralobular degeneration and focal necrosis, portal inflammation, and fibrosis. The evaluation of biliary morphological changes included the following parameters: bile duct damage, ductular proliferation, cholangiolitis, ductular metaplasia of periportal hepatocytes, phenotypic change from the hepatocytic to the biliary type of cytokeratin (HAEI) in liver cells. Each parameter was semi-quantitatively graded from 0 to 3. A cumulative biliary score was computed for each individual biopsy specimen by adding the individual biliary scores. Each specimen included more than six portal tracts for evaluation. In addition to routine stains such as HE, PAS with or without diastase, reticulin, collagen trichrome, bilirubin, immunostaining for cytokeratins (antibodies AE1, AE3, Ortho diagnostics) was also performed by standard Peroxidase Anti-Peroxidase technique. AE1 and AE3 are monoclonal antibodies (MoAb) directed against epithelial cytokeratins of high and low molecular weight either acid (AE1) or basic (AE3). The MoAb-AE1 recognizes the biliary type of cytokeratins (Moll's catalogue N19); the MoAb-AE3 recognizes the hepatocytic type ofcytokeratins (Moll's catalogue N7 and 8) (13). Hepatocytes shifting their phenotype from AE3 to AE1 positivity are indicated as HAE1 in this study.

Statistical analysis Modifications in biochemical parameters during follow-up were assessed using analyses of variance (ANOVA) and covariance (ANCOVA) (14,15). These statistical methods were considered appropriate for both the experimental design and the continuous variables. Moreover, ANCOVA also provides for possible biases from different basal values. Data normalization, a prerequisite for parametric tests, was obtained by transforming values into their relative log, which were then verified by the Kolmogorov-Smirnov test (15). Statistical analysis of the histological data was performed using the Mann-Whitney (15) (to evaluate the differences between the two treatments) and Wilcoxon tests (15) (to evaluate differences between

URSODEOXYCHOLIC ACID IN CHRONIC ACTIVE HEPATITIS TABLE 1 Initial characteristics of patients with chronic active hepatitis included in treatment Groups A (ursodeoxycholic acid) or B (placebo) Group

A

B

No. of patients Females % Age (mean years) Age range Jaundice Esophageal varices % HBsAg pos. % Anti-HCV pos. (Elisa) % AST (xULNV)* T bilirubin > 1.5 mg % Previous treatment No. with cirrhosis

18 33.3 45.1 26-61 none 5.5 16.6 72.2 4.2 none none 1

18 44.4 48.7 22-63 none 5.5 16.6 72.2 3.9 none none 1

317 in G r o u p B from cardiac problems. One patient (HCV positive) crossed to interferon therapy and three patients (one H b s A g positive, two H C V positive) spontaneously interrupted treatment due to low compliance. One further patient (HBsAg positive) in G r o u p A and two patients (one H B s A g positive, one HBsAg and HCV negative) in G r o u p B did not have a second liver biopsy. Thus the results of the biochemical parameters refer to 16 patients from G r o u p A and 12 from G r o u p B, while liver biopsy results refer to 15 patients from G r o u p A and 10 from G r o u p B.

Biochemical parameters

* ULNV=upper limits normal values.

basal and final values within each treatment). These tests are the most a p p r o p r i a t e for ordinal values. Linear regression analysis was performed to assess the relationship between biochemical parameters and histological activity.

Results Patient characteristics at entry were similar between the two groups (Table 1). All but one patient (HBsAg negative, hepatitis C virus (HCV) negative) from G r o u p A gave their formal consent. Seventeen patients received U D C A and 18 a placebo for 12 months. One o f the 17 patients in G r o u p A and six o f the 18 patients in G r o u p B did not complete the follow-up period and were thus considered drop-outs. The difference was not statistically significant. Treatment was interrupted in one case (HCV positive) in G r o u p A because o f pregnancy. Two patients (one H C V positive, one H B s A g and H C V negative) died

Baseline and follow-up values at 3, 6, 9 and 12 months are reported in Table 2. Results are expressed as times above upper limits o f normal values___SD. Baseline values did not differ between the two groups. Significant reductions in A L T , A S T and G G T were observed in all four follow-up controls in G r o u p A. In G r o u p B a significant decrease in A S T from baseline values was observed only at 9 and 12 months. Serum bilirubin prior to treatment was within normal limits and did not vary during treatment. Multifactorial covariance analysis showed that the reductions in A L T (F=4.61; p<0.05), A S T (F=11.37; p<0.001) and G G T (F=27.97; p<0.001) were significantly higher in G r o u p A. Alkaline phosphatase was normal in most cases before treatment and there were no variations from baseline values during follow-up in either group.

Biochemical remission Biochemical remission was not observed in either group during follow-up. W h e n the various enzymes were con-

TABLE 2 Serum enzyme levels before and during the follow-up Baseline

3 months

6 months

9 months

12 months

AST UDCA Placebo

(16) 4.14---2.23 (12) 4.07-+2.61

(16) 2.42---1.46" (12) 3.23-+1.49

(16) 2.48-+ 1.75" (12) 3.04-+1.56

(16) 2.74---1.93t (11) 2.66-+1.81t

(16) 1.74-+1.07" (12) 2.51_-&-l.29t

ALT UDCA Placebo

(16) 6.43-+5.71 (12) 4.18-+1.58

(16) 3.84-+2.38§ (12) 4.83-+2.15

(16) 3.54-+2.58* (12) 4.17-+2.12

(16) 3.85-+2.41§ (11) 3.48---I.51

(16) 3.26-+2.48* (12) 3.72-+1.72

GGT UDCA Placebo

(16) 2.02-+2.90 (12) 1.58-+0.64

(15) 0.90-+1.10" (12) 1.31-+0.62

(16) 0.82-+1.14" (12) 1.45-+1.04

(14) 0.97-+1.09" (11) 1.43-+1.60§

(15) 0.57-+0.29* (11) 1.25-+0.51

AP UDCA Placebo

(16) 0.85-+0.31 (12) 0.84-+0.35

(14) 0.88-+0.37 (12) 0.94-+0.30

(15) 0.87---0.34 (12) 0.85-+0.42

(15) 0.93-+0.30 (11) 0.89-+0.37

(15) 0.96-+0.39 (11) 1.06-+0.45§

* p<0.001 vs basal values; t p<0.01 vs basal values; § p<0.05 vs basal values. AST=aspartate aminotransferase. UDCA= ursodeoxycholic acid. Results are expressed as X ULNV-+ SD. Figures in parentheses refer to number of patients examined at each control.

318

A.F. ATTILI et al.

TABLE 3 Individual and cumulative hepatitis activity scores (means+-SD) before and after a 1-year treatment period with ursodeoxycholic acid (UDCA) (n= 15) or placebo (n= 10) in patients with chronic active hepatitis Baseline Periportal+-bridging necrosis UDCA 3.73--,1.10 Placebo 4.00---1.49

12 months

p

3.93---1.53 3.70+--1.83

ns ns

Intralobular degeneration and focal necrosis UDCA 2.00--.1.13 1.87---1.13 Placebo 2.00---1.05 2.10---1.20

ns ns

Portal inflammation UDCA 3.13--.0.35 Placebo 3.20"--0.42

3.40---0.83 3.40"--0.52

ns ns

2.40"--1.06 2.60"-,1.17

ns ns

11.60"-,3.27 11.80"-,3.33

ns ns

Liver histology As evaluated with the Knodell's scoring system, the hepatitis activity scores in the two groups before and after treatment are reported in Table 3. N o differences were found before or after treatment in either group, nor were there any significant variations from baseline values. The biliary scores in the two groups before and after treatment are reported in Table 4. A significant reduction in ductal metaplasia HAE1 and biliary scores was observed in U D C A - but not in placebo-treated patients.

12 months

p

0.47"--0.52 1.40---0.74

NS NS

Relationships between biochemical and histological rabies When all cases are considered together, linear regression analysis between biochemical parameters and histological activity indices (Knodell) failed to reveal any significant association before treatment. There was no relationship between variations (either absolute or percent) in biochemical parameters and the histological activity indices (Knodell). HAE1 and biliary scores were significantly and positively associated with baseline G G T serum levels (r=0.58; p=0.002 and r=0.49; p=0.01, respectively), while ductal metaplasia was positively associated with alkaline phosphatase (r=0.47; p=0.016). Percent variations of ductal metaplasia (r=0.65; p=0.001), CH (r=0.44; p=0.03), HAE1 (r=0.71; p<0.001) and biliary scores (r=0.67; p<.0.0001) after treatment were positively associated with G G T variations.

Ductular metaplasia of periportal hepatocytes UDCA 1.60-+0.63 0.87"--0.52 Placebo 1.20_+0.42 1.20"-.0.56

<0.055 NS

Discussion

Ductular proliferation UDCA 1.27"-,0.70 Placebo 1.80---0.63

1.20---0.56 1.60"-.0.70

NS NS

Cholangiolitis UDCA Placebo

0.30"--0.48 0.40-+0.52

NS NS

Fibrosis UDCA 2.07-+ 1.03 Placebo 2.20-+1.31 Cumulative Knodell's score UDCA 10.93"-.2.19 Placebo 11.40---2.63

TABLE 4 Individual and cumulative scores (means___SD) of biliary morphological changes before and after a 1-year treatment period with ursodeoxycholic acid (UDCA) (n= 15) or placebo (n= 10) in patients with chronic active hepatitis Baseline Bile duct damage UDCA 0.53"--0.64 Placebo 0.90+--0.74

0.47"--0.52 0.20+-0.41

Phenotypic changefi'om hepatocytic to biliary type of cytokerathTs hi liver cells UDCA 0.87_+0.74 0.53---0.49 <0.02 Placebo 0.50-+0.53 1.00-+0.94 NS Cumulative biliary score UDCA 4.73---1.98 Placebo 4.70"--1 . 3 4

3.07---1.62 5.30---2.98

<0.01 NS

sidered separately, cumulative remission of AST was observed in 25% and 16.6% of patients in Groups A and B, respectively. Cumulative remission of A L T occurred in only one patient in G r o u p A. Of the nine and ten patients in Groups A and B, respectively, with G G T levels above the upper limits o f normal values before treatment, cumulative remission of G G T was observed in eight (89%) and three (39%) cases.

Previous controlled and uncontrolled studies have reported biochemical improvements in patients with chronic active hepatitis following U D C A administration. This study confirms a significant reduction in serum transaminases and G G T during U D C A treatment, compared to placebo. Whatever the etiology of the disease, U D C A seems to have a beneficial effect on serum enzymes in chronic active hepatitis. The cytoprotective effect of U D C A may occur either by promoting the excretion of cytotoxic bile acids (CDCA, DCA), or by binding to hepatocyte or biliocyte membranes, thus preventing damage from cytotoxic bile acids (16). Recently Calmus & Poupon (17,18) suggested that U D C A may operate via an immunoregulatory mechanism, by suppressing H L A antigen expression on liver cells. The detailed morphological evaluation in the present study has clearly shown that the morphological expression of chronic hepatitis has two components: one strictly "hepatitic", the other "biliary". Despite the biochemical improvement following U D C A treatment, no significant histological variation was found using a stan-

URSODEOXYCHOLIC ACID IN CHRONIC ACTIVE HEPATITIS

dard scoring system (Knodell's) which investigates inflammatory activity. However, when a sensitive morphological system was used to evaluate more subtle signs of biliary damage (not included in Knodell's score system), significant histological variations were found in the U D C A group. The main cholestatic parameters modified in the U D C A group were ductal metaplasia, HAE1 and bile duct damage. The latter morphological changes correlated with the biochemical improvement, particularly the decrease in G G T . These results indicate that a score system which evaluates only histological activity is not appropriate for studies such as this one. In addition, it should be emphasized that a scoring system for the evaluation of cholestatic parameters c a n n o t be limited to routinely stained sections, since additional stains are required. M a n y in vitro and in vivo studies (19-23) have clearly demonstrated the cytoprotective effect of U D C A . The morphological counterpart (histological improvement) might well be reflected in the "biliary" c o m p o n e n t of the hepatitis process. Obviously, the existence of an additional morphological improvement in inflammation c a n n o t be ruled out. This might occur after the improvement in serum enzymes. T h a n k s to experience with socalled acute exacerbation in the course of CAH, when spontaneous (or interferon-induced) seroconversion from e to anti-e occurs, liver pathologists are now familiar with this fact. L o b u l a r activity lasts for several months after seroconversion and transaminase normalization. If this is true for U D C A , the histological follow-up should be extended beyond 12 months. In conclusion, the improvement in serum enzymes in patients with chronic active hepatitis may at least partly reflect morphological changes in the biliary tree component of the hepatitis process, changes which are similar to those which occur in patients with chronic cholestatic liver disease (primary biliary cirrhosis, primary sclerosing cholangitis) (24-30).

References 1. Ichida F. Clinical experience with ursodeoxycholic acid (S-Urso) for chronic hepatitis. Diagn Treat 1961; 36: 388. 2. Nakahara T, Yoshida H, Bando T, et al. Effect of Ursosan tablets on chronic hepatic disorders. Diagn Treat 1975; 50: 1070-3. 3. Wakahara T. Effectivenessof S-Urso tablet against chronic liver diseases. Diagn Treat 1975; 63: 1070. 4. Miyaji K. Effect of ursodeoxycholic acid on liver function of chronic liver disease. Jpn J Clin Exp Med 1976; 53: 1395. 5. Yamanaka M. Study of curative effect of ursodeoxycholic acid for chronic hepatitis by double-blind trial. Diagn Treat 1976; 64: 2150. 6. Leushner U, Leushner M, Sieratzki J, Kurtz W, Hubner K. Gallstone dissolution with ursodeoxycholic acid in patients with

319 chronic active hepatitis and two years follow-up. A pilot study. Dig Dis Sci 1985; 30: 642-9. 7. Attili AF, Angelico M, Cantafora A, Alvaro D, Capocaccia L. Bile acid-induced liver toxicity: relation to the hydrophobichydrophilic balance of bile acids. Med Hypotheses 1986; 19: 57-69. 8. Attili AF, Angelico M, Alvaro D, Marin M, De Santis A, Capocaccia L. Reduction in serum transaminases during taurine administration in patients with chronic active hepatitis. Gastroenterology 1984; 86: 1017. 9. Podda M, Ghezzi C, Battezzati PM, et al. Ursodeoxycholic acid for chronic liver disease. J Clin Gastroenterol 1988; 10: $25-31. 10. Ghezzi C, Zuin M, Battezzati PM, Podda M. Effects of ursodeoxycholate, taurine and ursodeoxycholate plus taurine on serum enzyme levels in patients with chronic hepatitis. Hepatology 1987; 7: 1110. 11. Ohya T. Long-term administration of ursodeoxycholic acid in patients with chronic hepatitis. A comparison of the effect of two doses on liver function. Jpn Pharmacol Ther 1985; 13: 271. 12. Knodell RG, Ishak KG, Black W, et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981; 1: 431-5. 13. Van Eyken P, Sciot R, Callea F, Van Der Steen K, Moerman P, Desmet VY. The development of the intrahepatic bile ducts in man: a keratin-immunohistochemical study. Hepatology 1988; 8: 1586-95. 14. Snedecor GW, Cochran WG. Statistical Methods. 6th Edn. Ames: Iowa State University Press, 1967. 15. Siegel S. Nonparametric Statistics for the Behavioural Sciences. New York: McGraw-Hill, 1956. 16. Hofmann AF, Barrett KE, Quist RG. Mechanisms of action of ursodeoxycholic acid in the treatment of chronic cholestatic hepatobiliary disease: experiments and speculations. Abstr Falk Symposium No. 53: I. Int Lugano Symposium on Biliary Physiology and Diseases: Strategies for the Treatment of Hepatobiliary Diseases. Lugano: June 15-17, 1989: 17. 17. Calmus Y, Gane P, Rouger P, Poupon R. Hepatic expression of HLA molecules in primary biliary cirrhosis: the effect of ursodeoxycholic acid. J Hepatol 1989; 9: S15. 18. Calmus Y, Gane P, Rouger P, Poupon R. Hepatic expression of class I and class II major histocompatibility complex molecules in primary biliary cirrhosis: effect of ursodeoxycholic acid. Hepatology 1990; 11: 12-5. 19. Sawkat Anwer M, Hondalus MK, Atkinson JM. Ursodeoxycholate-induced changes in hepatic Na ÷ H ÷ exchange and biliary HCO ~ excretion. Am J Physiol 1989; 257: 47-55. 20. Barnwell SG, Lowe PJ, Coleman R. Effect of taurochenodeoxycholate and tauroursodeoxycholate upon biliary output of phospholipids and plasma membrane enzymes, and the extent of cell damage, in isolated perfused rat livers. Biochem J 1983; 216: 56-60. 21. Miyazaki K, Nakayama F, Koga A. Effect of chenodeoxycholic and ursodeoxycholic acids on isolated adult human hepatocytes. Dig Dis Sci 1984; 29: 1123-30. 22. Kitani K, Ohta M, Kanai S. Tauroursodeoxycholate prevents biliary protein excretion induced by other bile salts in the rat. Am J Physiol 1985; 248: 168-78. 23. Scholmerich J, Becher MS, Schmidt K, et al. Influence of hydroxylation and conjugation of bile salts on their membranedamaging properties. Studies on isolated hepatocytes and lipid membrane vesicles. Hepatology 1984; 4: 661. 24. Poupon RE, Eschwege E, Poupon R and the UDCA-PBC Study Group. Ursodeoxycholic acid for the treatment of primary biliary cirrhosis. Interim analysis of a double-blind multicentre randomized trial. J Hepatol 1990; 11: 16-21. 25. Leuschner U, Fisher H, Kurtz W, Hubner K. Ursodeoxycholic

320 acid in primary biliary cirrhosis: results of a controlled doubleblind trial. Gastroenterology 1989; 97: 1268-74. 26. Hadziyannis SJ, Hadziyannis ES, Makris A. A randomised controlled trial of ursodeoxycholic acid in primary biliary cirrhosis. Hepatology 1989; 10:580 (abstract). 27. Bateson MC, Ross PE, Diffey BI. Ursodeoxycholic acid in primary biliary cirrhosis. Lancet 1989; i: 898-9. 28. Eisenburg Von J, Eder M, Spengler U, Berg PA, Caselmann

A. E ATTILI et al. W, Mannes AG, Muntau A. Ursodesoxycholsaure bei primar biliarer Zirrhose. Teil 2: prospektive Langzeitstudie an 21 Patienten. Fortschr Med 1988; 106: 428-31. 29. James OF. Ursodeoxycholic acid treatment for chronic cholestatic liver disease. J Hepatol 1990; I I: 5-8. 30. Chazouilleres O, Poupon R, Capron R, et al. Ursodeoxycholic acid for primary sclerosing cholangitis. J Hepatol 1990; 11: 120-3.