Clinical Significance of Serum Bilirubin Levels Under Ursodeoxycholic Acid Therapy in Patients With Primary Biliary Cirrhosis

Transcription

1 Clinical Significance of Serum Bilirubin Levels Under Ursodeoxycholic Acid Therapy in Patients With Primary Biliary Cirrhosis ANNE-MARIE BONNAND, 1 E. JENNY HEATHCOTE, 2 KEITH D. LINDOR, 3 AND RENÉE EUGÉNIE POUPON 1 We determined whether the normalization of serum bilirubin level (SBL) induced by ursodeoxycholic acid (UDCA) therapy was associated with an improved clinical outcome in patients with primary biliary cirrhosis (PBC). We estimated the prognostic values of SBL measured after 6 months of UDCA treatment for survival free of orthotopic liver transplantation (OLT). We used a database of 548 patients with PBC followed in three trials of UDCA. Among UDCA-treated patients, we compared survival free of OLT in patients with normalized SBL (I17 mol/l) with those who had persistently elevated SBL. Difference in survival was tested between UDCA-treated patients whose SBL normalized with treatment and placebo patients who had normal baseline SBL. We evaluated, in each treatment group, the prognostic value of 6-month SBL. Survival was estimated using the Kaplan-Meier method and compared by the Cox model. Survival free of OLT was significantly longer in patients who had normalized SBL (P F.0001; relative risk [RR]: 3.7, UDCA group). Survival free of OLT was not significantly different between UDCA patients with normalized SBL and placebo patients with a normal baseline SBL (P.69). For several cutoffs of 6-month SBL, RRs of OLT or death were similar in UDCA-treated and placebo patients: the RR of OLT or death associated with a 6-month SBL more than 30 mol/l was 6.0 for UDCA and 5.7 for placebo groups. In conclusion, normalization of SBL during therapy is associated with improved clinical outcome. SBL under UDCA therapy is a prognostic factor in PBC. SBL under UDCA therapy should be interpreted as in untreated patients. (HEPATOLOGY 1999;29:39-43.) Several trials of ursodeoxycholic acid (UDCA) in primary biliary cirrhosis (PBC) showed an improvement in biochemical parameters. 1-4 In addition, a combined analysis, pooling 548 patients involved in three large trials using a similar dose Abbreviations: UDCA, ursodeoxycholic acid; PBC, primary biliary cirrhosis; OLT, orthotopic liver transplantation; SBL, serum bilirubin level; RR, relative risk. From 1 INSERM Unit 21, Villejuif, France; 2 Division of Gastroenterology, The Toronto Hospital, Toronto, Ontario, Canada; and 3 Division of Gastroenterology, Mayo Clinic, Rochester MN. Received April 30, 1998; accepted August 13, Drs. Bonnand and Poupon s current address is: INSERM Unit 370, Paris, France. Presented in part at the 32nd Annual Meeting of the European Association for the Study of the Liver, April 9-12, 1997, London, UK. Address reprint requests to: Dr. R. E. Poupon, INSERM Unit 370, Faculté de Médecine, 156, rue de Vaugirard, Paris cedex 15, France. Fax: Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/0 of UDCA, has demonstrated that 4-year UDCA therapy increases survival free of orthotopic liver transplantation (OLT). 5 In untreated patients with PBC, elevated levels of serum bilirubin level (SBL) have been consistently found to be an independent predictor of a poor prognosis. 6-9 UDCA therapy substantially lowers SBL but has not been shown to consistently improve clinical and histological parameters. 1-4 Nevertheless, SBL is one of the criteria applied in the decision for referral to OLT. 10 Thus, an important unresolved question is whether the prognostic value of SBL is maintained during UDCA treatment: Does the SBL threshold currently assumed in the decision to perform transplantation remain valid in UDCA-treated patients? In our combined database of three long-term trials of UDCA, 5 biochemical measurements were available at several points during the 4-year period of follow-up. We undertook the present study to determine whether the normalization of SBL induced by UDCA therapy was associated with improvement in the clinical outcome of patients with PBC. Furthermore, we aimed to examine if the response to UDCA treatment in terms of SBL normalization was able to capture the effect of treatment on survival free of OLT. In addition, we compared the prognostic values of the SBL measured 6 months after randomization in UDCA and placebo groups. We therefore analyzed separately in each group the survival free of OLT according to the SBL at 6 months, and calculated in each group the risks associated with different levels of this parameter. PATIENTS AND METHODS Patient Population and Clinical Database. The data set for our combined analysis of three clinical trials of UDCA in PBC included 548 patients: 146 from the French trial, 180 from the American trial, and 222 from the Canadian trial. 5 Detailed patient characteristics and the study design are described elsewhere. 1-3 Patients were randomly assigned to receive UDCA (n 273) or a placebo (n 275). In the French and Canadian studies, patients were randomized to receive UDCA or an identical placebo for a 2-year period, and thereafter, patients were offered UDCA for 2 years. In the Mayo study, patients were randomized to receive UDCA or a placebo for up to 4 years. The two treatment groups were labeled as UDCA when originally randomized to UDCA and originally placebo when originally randomized to placebo regardless of the treatment they received after 2 years. At baseline, 146 patients assigned to UDCA had elevated SBL ( 17 µmol/l). The analysis of SBL normalization was performed in this subgroup of patients. Survival rates were calculated according to SBL status in this subgroup and compared with survival rates in the 275 placebo patients according to baseline SBL. In the second part, we focused on the group of 500 patients 39

2 40 BONNAND ET AL. HEPATOLOGY January 1999 (UDCA: n 245; originally placebo: n 255) for whom biological data, including SBL, were available after 6 months of treatment. Among the 48 patients (28 UDCA, 20 originally placebo) for whom these data were not available, 4 died and 6 underwent OLT between entry and 6 months, 15 dropped out of the trials before 6 months for other reasons, and 23 had an incomplete data set. Statistical Analysis. Patients bilirubin levels were regarded as normalized if the measurement was less than 17 µmol/l and continued on a consecutive measurement in the course of the trial. As SBL was measured at the examinations performed at 3-month intervals during the first 2 years and at 6-month intervals thereafter, the time at which normalization initially occurred was recorded as the midpoint between successive examinations (e.g., if the first SBL less than 17 µmol/l was measured at the 9-month examination, the time until normalization was recorded at 7.5 months). The proportion of patients with SBL normalization was estimated in the UDCA group by the Kaplan-Meier method. 11 The main end-point used in the present study was survival free of OLT. In the remainder of this article, the term survival has the meaning of survival free of OLT. Criteria for OLT generally included at least two of the following complications: an SBL above 150 µmol/l, hepatic encephalopathy, ascites, variceal bleeding, or intractable pruritus, the absence of intercurrent disease, and age under 65 years. Among UDCA-treated patients with elevated baseline SBL (n 146), we compared survival free of OLT in patients with normalized SBL and those whose SBL remained elevated despite treatment, using the Kaplan-Meier method. Patients whose SBL normalized should have had longer survival times than those who did not, because they lived at least until normalization. To avoid this bias, survival time was estimated as follows: first, for patients who had normalization, the time until normalization of SBL was subtracted from the observation time; then, data from all 146 patients were used for the Kaplan-Meier estimates of patients with elevated SBL, with normalization as a reason for censoring data. Thus, the two groups of UDCA-treated patients were made up as follows: The first group was composed of patients whose SBL normalized with a survival time starting from the date of normalization. The second group, with elevated SBL, comprised all the patients from baseline: patients whose SBL did not go on to normalize remained in this second group for all of follow-up; patients whose SBL went on to normalize were in this second group until the time of their normalization. The comparison between those two groups was performed by a Cox proportional hazard model adjusting for age and the study. Similarly, a survival analysis was performed among placebo patients to compare survival free of OLT between patients with a baseline SBL less 17 µmol/l and patients with baseline SBL more than 17 µmol/l. Then, to determine whether the clinical outcome of patients with SBL normalization was similar to that of untreated patients with normal SBL, we compared with a log rank test the survival free of OLT in the group of UDCA patients with normalized SBL to survival of placebo patients with normal baseline SBL. The same analysis was performed between UDCA patients with persistently elevated SBL and placebo patients with elevated baseline SBL. The results were confirmed by Cox models adjusting for age and the study. In the subset of 500 patients who had an SBL measurement at 6 months, we studied the chance of OLT or death according to 6-month SBL in each treatment group. These probabilities were estimated from survival curves using the Kaplan-Meier method and compared using the log rank test and a Cox regression model, adjusting for age at entry and for the study. Relative risks (RR) of OLT or death were calculated for several SBL cutoffs. Ratios of death and OLT in each treatment group were compared using the 2 test. All P values were two-sided, and P.05 was considered statistically significant. All analyses were performed on an intent-totreat basis. RESULTS Patient Characteristics. At entry into the studies, 273 patients were randomized to receive UDCA, and 275 a placebo. At the time of randomization, the two groups were well balanced for age ( vs years), gender (92.3% vs. 90.5% of women), and histological stage (56.6% vs. 57.4% stages III-IV) in the UDCA and original placebo groups, respectively. At 6 months, the main characteristics of the subset of 500 patients for whom a SBL test result was available were as follows: SBL: µmol/l vs µmol/l in the original UDCA and original placebo groups, respectively (P.001); alkaline phosphatase activity: 3.0_ 0.2 N vs N (N: upper limit of normal range; P.0001); aspartate transaminase activity: N vs N(P.0001); serum albumin level: g/l vs g/l (P.36). Normalization of SBL. In the UDCA group, 146 patients had a SBL more than 17 µmol/l at entry into the trials. Seventyone of these patients went on to normalize SBL under UDCA treatment. The median time to normalization was 1.5 months after the beginning of treatment (range, months) (Fig. 1). In this subgroup of 146 UDCA-treated patients, baseline characteristics of the patients with normalized SBL were compared with those of the patients who did not normalize this parameter during the follow-up (Table 1). It should be noted that age was not significantly different in these two groups of patients. In contrast, the disease was more severe as judged by liver test results (serum albumin and prothrombin time). The frequency of pruritus in patients who did not go on to normalize their SBL was higher than in the other group of patients. In this UDCA subgroup, cumulative survival (until OLT or death) was compared between patients whose SBL had normalized (n 71) and the group comprising all the patients with elevated SBL (n 146), those whose SBL went on to normalize until normalization, and those whose SBL remained elevated after the introduction of UDCA therapy (Fig. 2). Survival was significantly longer in patients whose SBL had normalized than in patients with elevated SBL (P.0001; RR: 3.5; 95% CI: ). These results were similar to those obtained when compar- FIG. 1. UDCA subgroup with elevated baseline SBL (n 146) Kaplan- Meier estimate of time to SBL normalization for the subgroup of UDCA patients whose baseline SBL was more than 17 µmol/l. Seventy-one of them went on to normalize SBL during follow-up.

3 HEPATOLOGY Vol. 29, No. 1, 1999 BONNAND ET AL. 41 TABLE 1. Original UDCA Group: Comparison of Baseline Characteristics of the Patients With Normalized SBL With Those Who Did Not Normalize This Parameter Parameter Normalized SBL (n 71) Not Normalized SBL (n 75) P Age (yr) * Laboratory findings Serum bilirubin (µmol/l) Alkaline phosphatase ( N) NS AST ( N) Serum albumin (g/l) Prothrombin time (s) IgM ( N) NS Clinic (present vs. absent) Pruritus 48/71 62/75.04 Fatigue 49/71 60/75 NS Hepatomegaly 13/45 17/39 NS Splenomegaly 10/45 12/39 NS Histological stage III-IV 45/70 56/72.08 NOTE. To convert values for SBL to mg/dl, divide by Abbreviations: N, upper limit of the normal range; AST, aspartate transaminase; IgM, immunoglobulin M, NS, not significant. *Mean SE. ing survival of original placebo patients with a baseline SBL of more than 17 µmol/l (n 152) to survival of placebo patients who had a normal baseline SBL (n 121) (P.0001; RR: 3.7; 95% CI: ) (Fig. 2). Thus, the normalization induced by UDCA therapy was associated with an improvement in survival free of OLT. Furthermore, the size of this improvement (RR) was similar to the difference in survival observed between placebo patients with normal or elevated baseline SBL. There was no statistically significant difference between survival of UDCA patients with normalized SBL and survival of placebo patients with normal baseline SBL (P.69). Patients who did not normalize SBL despite treatment had a survival rate similar to placebo patients with elevated baseline SBL (P.43). Given the SBL status, the survival free of OLT was not different in the UDCA and placebo patients; the biological response in terms of SBL normalization was able to capture the treatment effect on survival free of OLT. Predictive Value on Survival of 6-Month SBL During UDCA or Placebo Treatment. To evaluate the prognostic values of SBL after 6 months of UDCA therapy or placebo, we calculated the RRs of OLT or death for three SBL cutoffs: 20, 30, and 40 µmol/l (Table 2). Among the 245 UDCA patients having a 6-month SBL, 170 had a 6-month SBL less than 20 µmol/l, 202 had less than 30, and 220 had less than 40. Among the 255 placebo patients, 149 had a 6-month SBL less than 20 µmol/l, 188 had less than 30, and 206 had less than 40. The RRs of OLT or death were similar in the UDCA and the original placebo groups (Table 2). For the 30-µmol/L cutoff, the probability of survival according to 6-month SBL is shown in Fig. 3 in the original UDCA and placebo groups. In both groups, survival free of OLT was significantly longer in patients with a 6-month SBL less than 30 µmol/l than in those having a 6-month SBL more than 30 µmol/l (log rank test: UDCA P.0001, original placebo P.0001; Cox model, P.0001 in both analyses). In UDCA-treated patients, the risk of OLT or death was 6.0 times higher (95% CI: ) in patients with SBL more than 30 µmol/l compared with the others (Table 2). In the original placebo patients, the risk was 5.7 times higher (95% CI: ) for patients with SBL more than 30 µmol/l compared with the others. There were more deaths or OLT procedures in the placebo group, as previously reported. However, we wondered if in the UDCA group, the improvement in SBL might inappropriately delay OLT, leading to excess mortality. We compared the ratios of OLT or death in the two treatment groups. From 6 months to 4 years after inclusion in the trials (n 500), there were 21 deaths and 20 OLT in the UDCA group versus 28 and 29 in the original placebo group (P.99). In the subgroup of patients with a 6-month SBL higher than 30 µmol/l, there were 10 deaths and 11 OLT in the UDCA group compared with 14 deaths and 20 OLT in the original placebo group (P.85). FIG. 2. UDCA subgroup with elevated baseline SBL (n 146) (thick line): Cumulative survival free of OLT among patients having normalized SBL and those who still had elevated SBL. Survival was significantly longer in patients who had normalized SBL than in the others (P.0001; RR: 3.5; 95% CI: ). Original placebo group (thin line): Cumulative survival free of OLT among patients having a baseline SBL more than 17 µmol/l (n 152) and remaining placebo patients (n 121). Survival was significantly longer in patients who had a normal baseline SBL (P.0001; RR: 3.7; 95% CI: ). There was no statistically significant difference between survival of UDCA patients with normalized SBL and survival of placebo patients with normal baseline SBL (P.69). Patients who did not normalize SBL despite treatment had a survival rate similar to placebo patients with elevated baseline SBL (P.43). TABLE 2. Univariate Analysis of Prognostic Value of SBL Measured After 6-Month UDCA Treatment (n 245) or 6-Month Placebo Treatment (n 255) for Survival Free of Liver Transplantation Classes for SBL Treatment P RR (95% CI) 20 vs. 20 µmol/l Original UDCA ( ) Original placebo ( ) 30 vs. 30 µmol/l Original UDCA ( ) Original placebo ( ) 40 vs. 40 µmol/l Original UDCA ( ) Original placebo ( ) NOTE. To convert values for SBL to mg/dl, divide by Data were obtained by Cox regression analysis adjusting for age and study.

4 42 BONNAND ET AL. HEPATOLOGY January 1999 FIG. 3. Predictive value on survival of 6-month SBL during UDCA or placebo treatment. In the UDCA group (thick line) and the original placebo group (thin line), survival free of OLT was significantly lower in patients with a 6-month SBL more than 30 µmol/l than in those having a 6-month SBL less than 30 µmol/l (P.0001 in both treatment groups; RR: 6.0 UDCA group, 5.7 original placebo group). DISCUSSION Because UDCA has been consistently found to lower SBL, we aimed to evaluate the clinical benefit associated with UDCA-induced SBL normalization. Our results show that patients in whom SBL normalized during UDCA therapy had a better clinical outcome than patients whose measurements remained elevated despite treatment. The normalization of SBL under UDCA treatment had a prognostic implication for survival free of OLT. In addition, the size of the improvement in survival associated with SBL normalization was similar to the survival increment associated with normal baseline SBL in placebo patients. The estimated survival rate of UDCA patients who responded to treatment in terms of SBL normalization was not different from the survival rate of placebo patients with normal baseline SBL. Patients who did not normalize SBL despite treatment had a survival rate similar to placebo patients with elevated baseline SBL. In other words, given SBL status, survival free of OLT was independent of the treatment group. The treatment effect on survival free of OLT previously reported 5 was captured by the biological response. These results show that the response to UDCA treatment in terms of SBL normalization fulfills for survival free of OLT the two criteria defined by Prentice 12 for a surrogate marker: 1) the normalization of SBL under UDCA treatment has a prognostic implication for survival free of OLT; 2) the survival free of OLT is independent of treatment, conditional on the SBL status. In clinical practice, this marker is easy to use. Because the SBL normalization induced by UDCA treatment occurs very early after the beginning of therapy, this marker could be useful in identifying patients who are unlikely to respond to UDCA treatment alone. A subset of patients who have persistent abnormalities under UDCA may progress toward cirrhosis and thus may require adjuvant therapy. So far, three classes of drugs have been tested in combination with UDCA: colchicine, methotrexate, and corticosteroids Briefly, 1) the benefit of colchicine appeared to be marginal or absent in UDCA-treated patients; 2) methotrexate had little further benefit in combination with UDCA; and 3) the combination of corticosteroids with UDCA showed improvement in biochemical results with or without improvement in liver histology. Because most of them are preliminary studies (open, pilot, or short-term follow-up), these results must be confirmed in long-term controlled trials. It is noteworthy that there was no difference in the age of patients normalizing SBL compared with the remaining patients. However, the presence of pruritus at entry into the studies was significantly more frequent in patients who did not go on to normalize SBL under UDCA treatment. The absence of normalization was also associated with significantly higher baseline aspartate transaminase, SBL, and prothrombin time and lower serum albumin levels. Our results are in keeping with those of Jorgensen et al. 25 defining the characteristics of patients with complete biochemical response to UDCA. In the original UDCA-treated group, normalization occurred within the first 24-month period, with a median time of 1.5 months. Furthermore, we sought to complement these findings on SBL normalization by studying the survival of patients with PBC treated with UDCA according to several SBL classes above the upper limit of the normal range. We determined which cutoffs of SBL measured during UDCA treatment had a predictive value for the patients outcome. We therefore focused on SBL values at the 6-month measurement. This point was chosen for two reasons: first, because in UDCAtreated patients most SBL normalizations occurred during the first 6-month period of therapy, we assumed this period was long enough to detect a possible UDCA-induced decrease in SBL; second, our database was comprehensive enough for the values of this parameter at 6 months to avoid selection bias. The SBL cutoffs studied were: 20, 30, and 40 µmol/l. As in the original placebo group, for all these cutoffs, a higher SBL in those receiving UDCA therapy was associated with shorter survival without transplantation. In addition, for a given SBL cutoff, the RR of OLT or death was similar in UDCA-treated patients and original placebo patients. Thus, 6-month SBL remains a prognostic factor of OLT or death. These results are in accordance with previous findings 26,27 that the Mayo score is still valid in UDCA-treated patients. More elevated SBL cutoffs than those tested cannot be used, because there were not enough events or patients. We determined that there was no OLT deficit in the UDCA group. As shown previously, the overall number of events is lower in the UDCA group than in the original placebo group; however, the ratios of OLT and death are similar in the two treatment groups. Similar results were observed in the subset of patients with 6-month SBL higher than 30 µmol/l, i.e., a difference in the number of events, and an absence of differences in the ratios of OLT and death. Thus, the criticism that UDCA treatment could mask the severity of the disease and inappropriately delay referral for OLT was unwarranted. 28 In conclusion, the UDCA-induced normalization of SBL is associated with clinical benefit in terms of survival free of OLT. Because this biological response can be measured early in the course of therapy, it can help to identify nonresponders to UDCA therapy. In addition, the association between a given value of SBL and the risk of OLT or death is the same in patients treated with UDCA as in original placebo patients. From a clinical standpoint, this means that in UDCA-treated

5 HEPATOLOGY Vol. 29, No. 1, 1999 BONNAND ET AL. 43 patients with PBC, SBL should be interpreted as for untreated patients. REFERENCES 1. Poupon RE, Balkau B, Eschwège E, Poupon R, and the UDCA-PBC Study Group. A multicenter, controlled trial of ursodiol for the treatment of primary biliary cirrhosis. N Engl J Med 1991;324: Lindor KD, Dickson ER, Baldus WP, Jorgensen RA, Ludwig J, Murtaugh P, Harrison JM, et al. Ursodeoxycholic acid in the treatment of primary biliary cirrhosis. Gastroenterology 1994;106: Heathcote EJ, Cauch-Dudek K, Walker V, Bailey RJ, Blendis LM, Ghent CN, Michieletti P, et al. The Canadian multicenter double-blind randomized controlled trial of ursodeoxycholic acid in primary biliary cirrhosis. HEPATOLOGY 1994;19: Combes B. Carithers R, Maddrey WC, Li D, McDonald MF, Wheeler DE, Eidenbdrot EH, et al. A randomized, double-blind, placebo-controlled trial of ursodeoxycholate in primary biliary cirrhosis. HEPATOLOGY 1995;22: Poupon RE, Lindor KD, Cauch-Dudek K, Dickson ER, Poupon R, Heathcote EJ. Combined analysis of randomized controlled trials of ursodeoxycholic acid in primary biliary cirrhosis. Gastroenterology 1997;113: Shapiro JM, Smith H, Schaffner F. Serum bilirubin: a prognostic factor in primary biliary cirrhosis. Gut 1979;20: Roll J, Boyer JL, Barry D, Klatskin G. The prognostic importance of clinical and histological features in asymptomatic and symptomatic primary biliary cirrhosis. N Engl J Med 1983;308: Christensen E, Neuberger J, Crowe JP, Altman DG, Popper H, Portmann B, Doniach D, et al. Beneficial effect of azathioprine and prediction of prognosis in primary biliary cirrhosis. Final results of an international trial. Gastroenterology 1985;89: Dickson ER, Grambsch PM, Fleming TR, Fisher LD, Langworthy A. Prognosis in primary biliary cirrhosis: model for decision making. HEPATOLOGY 1989;10: Consensus statement on indications for liver transplantation: Paris, June 22-23, HEPATOLOGY 1994;20:63S-68S. 11. Kaplan EL, Meier P. Non-parametric estimation from incomplete samples. J Am Stat Assoc 1958;53: Prentice RL. Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med 1989;8: Raedsch R, Stiehl A, Walker S, Theilman L, Kommerell B. Effects of ursodeoxycholic acid and ursodeoxycholic acid plus colchicine in primary biliary cirrhosis: a double-blind pilot study. In: Paumgartner G, Stiehl A, Gerok W, eds. Bile Acids as Therapeutic Agents. From Basic Science to Clinical Practice. Dordrecht: Kluwer Academic Publishers, 1991: Almasio P, Provenzano G, Battezzati PM, Podda M, Todros L, Rosina F, Saccocio G, et al. The Italian multi-centre randomized controlled trial of ursodeoxycholic acid vs ursodeoxycholic acid plus colchicine in symptomatic primary biliary cirrhosis [Abstract]. HEPATOLOGY 1994;20: Ikeda T, Tozuka S, Noguchi O, Kobayashi F, Sakamoto S, Marumo F, Sato C. Effects of additional administration of colchicine in ursodeoxycholic acid treated patients with primary biliary cirrhosis: a prospective randomized study. J Hepatol 1996;24: Poupon RE, Huet PM, Poupon R, Bonnand A-M, Tran van Nhieu J, Zafrani ES, and the UDCA-PBC Study Group. A randomized trial comparing colchicine and ursodeoxycholic acid combination to ursodeoxycholic acid in primary biliary cirrhosis. HEPATOLOGY 1994;24; Buscher H-P, Zietzschmann Y, Gerok W. Positive responses to methotrexate and ursodeoxycholic acid in patients with primary biliary cirrhosis responding insufficiently to ursodeoxycholic acid alone. J Hepatol 1993;18: Lindor KD, Dickson ER, Jorgensen RA, Anderson ML, Wiesner RH, Gores GJ, Lange SM, et al. The combination of ursodeoxycholic acid and methotrexate for patients with primary biliary cirrhosis: the results of a pilot study. HEPATOLOGY 1995;22: Van Steenbergen W, Sciot R, Van Eyken P, Desmet V, Fevery J. Combined treatment with methotrexate and ursodeoxycholic acid in non-cirrhotic primary biliary cirrhosis. Acta Clin Belg 1996;51: Gonzales-Koch A, Brahm J, Antezana C, Smok G, Cumsille MA. The combination of ursodeoxycholic acid and methotrexate for primary biliary cirrhosis is not better than ursodeoxycholic acid alone. J Hepatol 1997;27: Wolfhagen FHJ, van Buuren HR, Schalm SW. Combined treatment with ursodeoxycholic acid and prednisolone in primary biliary cirrhosis. Neth J Med 1994;44: Leuschner M, Gültdütuna S, You T, Hübner K, Bhatti S, Leuschner U. Ursodeoxycholic acid and prednisolone versus ursodeoxycholic acid and placebo in the treatment of early stages of primary biliary cirrhosis. J Hepatol 1996;25: Van Hoogstraten HJF, Wolfhagen FHJ, Van Berge Henegouwen GP, Schalm SW, ten Kate FJW, Hop WCJ, Van Buuren HR, et al. Combined bile acid immunosuppressive therapy for primary biliary cirrhosis. Results of a 1-year multicentre, placebo controlled study [Abstract]. HEPATOLOGY 1996;24: Leuschner U, Maier KP, Güldütuna S, Parté-Peterhans S, Leuschner M. Ursodeoxycholic acid in combination with prednisolone or budenoside in the therapy of primary biliary cirrhosis. In: Paumgartner G, Stiehl A, Gerok W, eds. Bile Acids in Hepatobiliary Diseases. Basic Research and Clinical Application. Dordrecht: Kluwer Academic Publishers, 1997: Jorgensen RA, Dickson ER, Hofmann AF, Rossi S, Lindor KD. Characterization of patients with a complete biochemical response to ursodeoxycholic acid. Gut 1995;36: Kilmurry MR, Heathcote EJ, Cauch-Dudek K, O Rourke K, Bailey RJ, Blendis LM, Ghent CN, et al. Is the Mayo model for predicting survival useful after introduction of ursodeoxycholic acid treatment for primary biliary cirrhosis? HEPATOLOGY 1996;23: Lindor KD, Therneau TM, Hermans JE, Dickson ER. Mayo risk score accurately predicts patient outcome with ursodeoxycholic acid treatment of primary biliary cirrhosis [Abstract]. HEPATOLOGY 1996;24: Jones DEJ, James OFW, Bassendine MF. Ursodeoxycholic acid therapy in primary biliary cirrhosis. [Comments]. HEPATOLOGY 1995;21: