Autoimmune, Cholestatic and Biliary Disease
Bile acid changes after high-dose ursodeoxycholic acid treatment in primary sclerosing cholangitis: Relation to disease progression†
Article first published online: 1 MAR 2010
Copyright © 2010 American Association for the Study of Liver Diseases
Volume 52, Issue 1, pages 197–203, July 2010
How to Cite
Sinakos, E., Marschall, H.-U., Kowdley, K. V., Befeler, A., Keach, J. and Lindor, K. (2010), Bile acid changes after high-dose ursodeoxycholic acid treatment in primary sclerosing cholangitis: Relation to disease progression. Hepatology, 52: 197–203. doi: 10.1002/hep.23631
Potential conflict of interest: Dr. Lindor received grants from Axcan.
- Issue published online: 23 JUN 2010
- Article first published online: 1 MAR 2010
- Manuscript Accepted: 17 FEB 2010
- Manuscript Received: 23 OCT 2009
- National Institute of Diabetes and Digestive and Kidney Diseases. Grant Number: DK56924
High-dose (28-30 mg/kg/day) ursodeoxycholic acid (UDCA) treatment improves serum liver tests in patients with primary sclerosing cholangitis (PSC) but does not improve survival and is associated with increased rates of serious adverse events. The mechanism for the latter undesired effect remains unclear. High-dose UDCA could result in the production of hepatotoxic bile acids, such as lithocholic acid (LCA), because of limited small bowel absorption of UDCA and conversion of UDCA by bacteria in the colon. We determined the serum bile acid composition in 56 patients with PSC previously enrolled in a randomized, double-blind controlled trial of high-dose UDCA versus placebo. Samples for analysis were obtained at the baseline and at the end of treatment. The mean changes in the UDCA level (16.86 versus 0.05 μmol/L) and total bile acid level (17.21 versus −0.55 μmol/L) were significantly higher in the UDCA group (n = 29) versus the placebo group (n = 27) when pretreatment levels were compared (P < 0.0001). LCA was also markedly increased (0.22 versus 0.01 μmol/L) in the UDCA group compared to the placebo group (P = 0.001). No significant changes were detected for cholic acid, deoxycholic acid, or chenodeoxycholic acid. Patients (n = 9) in the UDCA group who reached clinical endpoints of disease progression (the development of cirrhosis, varices, liver transplantation, or death) tended to have greater increases in their posttreatment total bile acid levels (34.99 versus 9.21 μmol/L, P < 0.08) in comparison with those who did not. Conclusion: High-dose UDCA treatment in PSC patients results in marked UDCA enrichment and significant expansion of the total serum bile acid pool, including LCA. HEPATOLOGY 2010
Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease characterized by inflammation and destruction of the extrahepatic and/or intrahepatic bile ducts, and it results in biliary cirrhosis, the need for liver transplantation, and reduced life expectancy.1 Up to now, there have been no reports of a medical therapy able to halt disease progression. Ursodeoxycholic acid (UDCA), initially tested at a dose of 13 to 15 mg/kg/day, showed some beneficial effects in patients with PSC as measured by liver biochemistry.2 Subsequent studies with higher drug doses showed even more favorable outcomes.3-6 However, a recently published randomized, double-blind controlled trial of high-dose (28-30 mg/kg/day) UDCA versus placebo failed to demonstrate improvement in survival, and UDCA was associated with increased rates of serious adverse events.7
The mechanism for this unexpected detrimental effect remains unclear. It has been postulated that high-dose UDCA treatment allows unabsorbed drug to enter the colon and be modified into hydrophobic, hepatotoxic bile acids, such as lithocholic acid (LCA).8-10 LCA is hepatotoxic in animal models and leads to segmental bile duct obstruction, destructive cholangitis, and periductal fibrosis.11, 12 Nonetheless, a recent study testing the effects of various, escalating UDCA doses on biliary composition showed only minimal changes in all bile acids except UDCA, which was proportionally enriched.13
The aim of our study was to determine the serum bile acid composition after high-dose UDCA treatment during a randomized, double-blind controlled trial and to correlate the changes in bile acid levels with clinical outcomes.
Patients and Methods
Patients were entered into the present study according to the criteria followed for our randomized, double-blind controlled trial of high-dose UDCA versus placebo.7 Difficulties related to the multicenter nature of the study and the long enrollment period did not allow all of the initial study patients to be analyzed with respect to the bile acid composition. The study was approved by the institutional review boards at each site.
A PSC diagnosis was based on the following criteria: (1) chronic cholestatic disease of at least 6 months' duration; (2) a serum alkaline phosphatase level at least 1.5 times the upper limit of normal (ULN); (3) retrograde, operative, percutaneous, or magnetic resonance cholangiography findings consistent with PSC within 1 year of study entry; and (4) a liver biopsy sample in the previous year that was compatible with the diagnosis of PSC and was available for review.
Patients were excluded if they had any of the following: (1) coexisting conditions that would limit their life expectancy to less than 2 years; (2) an inability to provide consent; (3) treatment with UDCA, corticosteroids, or immunosuppressives in the 3 months prior to study entry; (4) end-stage liver disease as determined by clinical and laboratory parameters; (5) previous intraductal stones or operations on the biliary tree other than cholecystectomy, and (6) findings of another liver disease such as hepatitis B or C, hemochromatosis, Wilson disease, or primary biliary cirrhosis.
Patients received UDCA at a dose of 28 to 30 mg/kg/day (Axcan Pharma, Mont-St. Hiliare, Canada) in divided doses given with meals or an identical placebo.
Serum samples were obtained at entry into the study and at the end of treatment. All available paired samples were retrieved and used for analysis of the bile acid composition. Disease progression to cirrhosis, development of varices, cholangiocarcinoma, liver transplantation, and death during the trial were considered clinical endpoints.
Serum bile acids were analyzed qualitatively by conventional gas chromatography-mass spectrometry (GCMS) and quantitatively by isotope-dilution GCMS, as described previously, with the following modification: alkaline hydrolysis with 2 M sodium hydroxide in 90% (vol/vol) ethanol (1 hour at 67°C) was performed instead of enzymatic hydrolysis with cholylglycine hydrolase.14 Deuterium-labeled LCA, deoxycholic acid (DCA), chenodeoxycholic acid (CDCA), cholic acid (CA), and UDCA as internal standards were obtained from CDN Isotopes (Pointe-Claire, Canada).
Baseline characteristics were calculated as medians and ranges for continuous variables. The number and percent in each group were tabulated for categorical variables. Bile acid concentrations were calculated as means and standard deviations. The chi-square test of independence was used to determine statistical significance for categorical data. For the continuous variables, the Wilcoxon rank sum test was used. Baseline characteristics with a significance of P ≤ 0.2 were entered into a multivariate model of multiple linear regression analysis to explore possible correlations per bile acid.
From 2001 to 2005, 150 patients with PSC were entered into the study. Serum for bile acid analysis was available at the baseline and at the end of the study (mean treatment duration = 2.38 ± 0.56 years) for 56 of these patients. The baseline characteristics of this subset of patients versus the remaining patients (n = 94) are shown in Table 1. Patients analyzed for their bile acid composition were younger (41.8 versus 49.3 years), were more likely to have concomitant inflammatory bowel disease (93% versus 68%), and had a lower Mayo risk score (0.015 versus 0.58). The baseline characteristics of this cohort of patients by treatment were comparable as well (Table 2).
|Characteristic||Bile Acid Analysis (n = 56)||No Bile Acid Analysis (n = 94)||P Value|
|Age, years||41.8 (17.8-75.5)||49.3 (20.3-73.6)||0.0017|
|Duration of disease, years||0.69 (0.04-7.5)||1.62 (0.005-49.5)||0.09|
|Female sex, n (%)||23 (41)||41 (44)||0.76|
|Colitis, n (%)||52 (93)||63 (68)||0.0004|
|Varices, n (%)||8 (14)||18 (19)||0.36|
|Histological stage, n (%)|
|I||15 (26)||35 (38)||0.09|
|II||20 (35)||20 (22)||0.07|
|III||13 (23)||24 (26)||0.54|
|IV||10 (18)||13 (14)||0.47|
|Alkaline phosphatase, U/L*||4.0 (1.2-12.5)||3.3 (0.52-26.9)||0.25|
|Aspartate aminotransferase, U/L*||1.6 (0.6-4)||1.9 (0.5-8.0)||0.12|
|Bilirubin, mg/dL*||0.9 (0.3-4.5)||1 (0.3-7.2)||0.18|
|Mayo risk score||0.015 (−1.05 to 1.8)||0.58 (−1.4 to 2.4)||0.0015|
|Characteristic||UDCA (n = 29)||Placebo (n = 27)||P Value|
|Age, years||43.4 (20.4-75.5)||38.3 (17.8-63.1)||0.09|
|Duration of disease, years||0.78 (0.06-7.5)||0.69 (0.04-6.3)||1|
|Female sex, n (%)||14 (48)||9 (33)||0.25|
|Colitis, n (%)||27 (93)||25 (93)||0.94|
|Colectomy, n (%)||7 (24)||5 (19)||0.60|
|Alkaline phosphatase, U/L*||3.8 (1.2-12.5)||4.4 (1.2-8.1)||0.84|
|Aspartate aminotransferase, U/L*||1.8 (0.7-4)||1.5 (0.6-4)||0.64|
|Bilirubin, mg/dL*||0.7 (0.3-2)||1 (0.3-4.5)||0.09|
|Mayo risk score||0.03 (−1.05 to 1.8)||−0.05 (−1.04 to 1.2)||0.50|
Baseline Bile Acid Composition.
At the baseline, the bile acid pool consisted of CA (32%), LCA (13%), DCA (21%), UDCA (11%), and CDCA (23%). GCMS spectra indicating the prevalence of significant amounts (>0.05 μmol/L) of uncommon bile acids were not observed. Only traces of hydroxylation products of UDCA were occasionally seen.15 The GCMS systems that were used did not separate UDCA from isoUDCA.
Patients who had undergone colectomy (n = 12) had significantly lower levels of DCA (P < 0.0001), whereas patients with a baseline alkaline phosphatase level ≥ 4 × ULN (n = 28) had lower levels of CA (P = 0.03; Table 3). Disease severity at entry, as assessed by the total bilirubin level, Mayo risk score, and histological stage, did not seem to considerably affect the baseline bile acid composition, although patients with a baseline total bilirubin level ≥ 0.9 mg/dL had higher values of CA (P = 0.05). In a multivariate analysis model, the only significant relationship that was revealed was between colectomy (P = 0.001), a baseline alkaline phosphatase level ≥ 4 × ULN (P = 0.05), and low levels of DCA.
|DCA (μmol/L)||CA (μmol/L)||CDCA (μmol/L)||UDCA (μmol/L)||LCA (μmol/L)|
|Colectomy (n = 12)||0.07 ± 0.05||0.54 ± 0.55||0.32 ± 0.25||0.07 ± 0.09||0.08 ± 0.07|
|No colectomy (n = 44)||0.27 ± 0.22||0.76 ± 2.44||0.40 ± 0.88||0.13 ± 0.14||0.11 ± 0.07|
|Alkaline phosphatase < 4 × ULN (n = 28)||0.27 ± 0.25||0.95 ± 3.06||0.46 ± 1.09||0.10 ± 0.10||0.12 ± 0.06|
|Alkaline phosphatase ≥ 4 × ULN (n = 28)||0.18 ± 0.16||0.47 ± 0.42||0.30 ± 0.22||0.14 ± 0.15||0.09 ± 0.08|
|Total bilirubin < 0.9 mg/dL (n = 27)||0.25 ± 0.22||0.36 ± 0.46||0.26 ± 0.24||0.08 ± 0.06||0.10 ± 0.08|
|Total bilirubin ≥ 0.9 mg/dL (n = 29)||0.20 ± 0.21||1.03 ± 2.98||0.49 ± 1.06||0.15 ± 0.16||0.11 ± 0.06|
Changes in Bile Acid Levels After Treatment.
Figure 1 shows the posttreatment percentage of each bile acid per treatment group. No significant changes between treatment groups were detected for CA, DCA, or CDCA. UDCA was significantly increased (16.86 versus 0.05 μmol/L, P < 0.0001), and the total bile acid pool was significantly expanded (17.21 versus −0.55 μmol/L, P < 0.0001) in the UDCA group versus the placebo group. LCA was also markedly increased in the UDCA group versus the placebo group (0.22 versus 0.01 μmol/L, P = 0.001). The change in LCA levels after UDCA treatment seemed to positively correlate with the change in UDCA levels (P = 0.19).
Correlation with Changes in Liver Tests and Clinical Outcome.
The UDCA and LCA enrichment did not show any significant relationship with the changes in the values of liver tests (levels of alkaline phosphatase, aspartate aminotransferase, and bilirubin and Mayo risk scores; data not shown). However, female and older patients were more likely to have a greater increase in their LCA value after UDCA treatment (Table 4). Patients who had undergone colectomy (n = 7) tended to have less LCA increase after treatment than those who had not undergone colectomy (Fig. 2). However, patients who had undergone colectomy did not have worse outcomes, regardless of the treatment group.
|Characteristic||Minimal ΔLCA (n = 14): 0.002 μmol/L (−0.14 to 0.07 μmol/L)||Marked ΔLCA (n = 15): 0.32 μmol/L (0.13-1.28 μmol/L)||P Value|
|Age, years||42.4 (20.4-49.8)||48.7 (30.7-75.5)||0.05|
|Female sex, n (%)||4 (29)||10 (67)||0.04|
|Body mass index, kg/m2||27.4 (23.8-33.1)||26.5 (20.6-38.8)||0.71|
|Colectomy, n (%)||4 (29)||3 (20)||0.58|
|Histological stage ≥ 2, n (%)||11 (79)||10 (67)||0.66|
|Alkaline phosphatase, U/L*||3.5 (1.2-12.1)||4.7 (1.41-12.5)||0.34|
|Aspartate aminotransferase, U/L*||1.6 (0.8-3.8)||2.2 (0.7-4)||0.94|
|Bilirubin, mg/dL*||0.7 (0.3-2)||0.8 (0.4-2)||0.74|
|Mayo risk score||−0.24 (−0.67 to 1.8)||0.36 (−1.05 to 1.5)||0.31|
Patients in the UDCA group who reached clinical endpoints during therapy (n = 9) tended to have higher increases in their LCA and total bile acid levels in comparison with those who did not (Fig. 3). The increase in total bile acids was almost entirely due to enrichment with UDCA. Table 5 summarizes the range of bile acid changes in these patients. The changes were similar in all patients except for one patient (patient 5), and this possibly indicated noncompliance.
|Patient No.||ΔDCA (μmol/L)||ΔCA (μmol/L)||ΔCDCA (μmol/L)||ΔUDCA (μmol/L)||ΔLCA (μmol/L)||ΔtBA (μmol/L)|
UDCA has shown some beneficial effects in patients with PSC.2 The inability to demonstrate slowing of disease progression has resulted over the last decade in several studies designed to explore the effectiveness of different UDCA doses.3-6 In the most recent study, high-dose (28-30 mg/kg/day) UDCA treatment was associated with increased rates of serious adverse events without any obvious explanation.7 Modification of the bile acid composition has been speculated to potentially underlie the effects of the drug.
In our present study, we investigated the serum bile acid composition in PSC patients under high-dose UDCA treatment. At the baseline, the primary bile acids CA and CDCA predominated. Disease severity, as assessed with biochemical markers, did not seem to considerably affect the bile acid composition. At the end of treatment, we found a significant expansion of the total serum bile acid pool and marked UDCA and LCA enrichment in the UDCA-treated patients versus the placebo group when pretreatment levels were compared. Additionally, we found that the increases in serum bile acid levels seemed to correlate with worse outcomes because the subset of patients that reached the clinical endpoints of disease progression during UDCA therapy tended to have higher bile acid levels.
The proposed UDCA mechanisms of action in hepatobiliary disorders include expansion of the hydrophilic bile acid pool and hypercholeresis.16 Early studies in gallstone patients showed that UDCA administration could modify the composition of circulating bile acids and lead to UDCA predominance.17, 18 Multiple subsequent studies, most of them in patients with primary biliary cirrhosis, have verified this modification and have generally revealed an overall expansion of the total bile acid pool.8, 9, 13, 14, 19-22 Rost et al.13 in a study of biliary bile acid composition in patients with PSC specifically indicated that UDCA enrichment was augmented (43.1% ± 0.3% to 58.6% ± 2.3% of total bile acids) parallel to an escalating dose of UDCA (10-13 to 22-25 mg/kg/day) and reached a plateau at the highest dose. This enrichment did not further increase beyond that dose.13
In our study, the mean changes in UDCA and total bile acid levels post-treatment were significantly higher in the UDCA group compared to the placebo group (P < 0.0001). The mean posttreatment UDCA percentage in the bile acid pool was 74%, far higher than that ever reported in any other study, and this implies that enrichment is increased proportionally to the dose. Nonetheless, this high enrichment did not correspond to a better outcome. Therefore, further investigation of biliary enrichment has to be performed, especially with respect to clinical outcome.
Changes in the levels of other bile acids under UDCA treatment have been generally considered to follow an inverse relationship between the increase in UDCA levels and the decrease in CA, CDCA, DCA, and LCA levels. Results, however, are not homogeneous, and the changes are not significant in the majority of cases.13, 21, 22 In PSC patients, one study has suggested that LCA levels are not increased, even after high-dose UDCA treatment.13 Nonetheless, the antibiotics administrated in that study during the endoscopic retrograde cholangiography procedure used to obtain samples for bile acid analysis might have interfered with the results.
Nonsignificant changes between the two treatment groups for CDCA, CA, and DCA levels were observed in our study after treatment. Levels of CDCA, CA, and DCA were slightly decreased in the placebo group, whereas in the UDCA group, CA showed a tendency to decrease, and DCA and CDCA were slightly increased. However, we found that LCA was clearly increased after treatment in the UDCA group versus the placebo group (P = 0.001).
An increase in the level of LCA could possibly represent a result of the high-dose UDCA treatment because LCA is mainly produced by bacterial 7-dehydroxylation of unabsorbed bile acid that passes into the colon.8-10, 23 UDCA absorption has been generally shown to be slow and incomplete.24-26 Moreover, it is inversely related to the severity of cholestasis.25 PSC patients are expected to experience various levels of cholestasis; in our study, however, total bilirubin, as a marker of cholestasis, was not significantly elevated. However, the dose of UDCA given was among the highest ever tried in PSC patients. LCA levels could have been influenced by the surgical removal of the colon.27 Nevertheless, in our subset of patients who undergone colectomy, no significant differences from patients with an intact colon were demonstrated; this may be due to the limited number of patients. In addition, five of seven patients who had undergone colectomy in the UDCA group underwent an ileal pouch procedure, which may have potentially interfered with the amount of LCA production.
Under normal conditions, bile is already relatively toxic, but actual liver toxicity is prevented by various mechanisms, including maintenance of the appropriate bile composition and normal bile flow.28 High levels of LCA disrupt this equilibrium because this bile acid is toxic per se and highly hydrophobic. In addition, LCA has been proven to promote bile duct injury in animal models through obstruction by LCA crystals and finally result in destructive cholangitis.12 In our study, LCA levels tended to be higher in patients in the UDCA group who reached clinical endpoints of disease progression versus those who did not. This relationship did not reach statistical significance, but this may be due to the small number of patients that reached a clinical endpoint. However, we think that our data are currently not solid enough to support the hypothesis that the worse outcome seen in this subset of patients can be explained solely by an increase in LCA levels.
The link of LCA action to cholangitis that is implied by our findings would certainly be exciting. However, we suggest that other potential explanations for the paradoxical effect that UDCA has in some patients treated with high doses also have to be investigated before final conclusions are drawn on UDCA mechanisms of action in PSC patients. Bile infarct aggravation due to increased bile flow and biliary pressure in the setting of biliary obstruction and modulation of apoptosis due to activated stellate cell life prolongation could present alternate mechanisms.29, 30 In our study, bile acid levels tended to be higher in patients in the UDCA group who reached clinical endpoints of disease progression versus those who did not. These changes could be useful as a way of assessing disease severity and following the disease course. Further studies focused on the relationships between bile acid levels and markers of disease progression are warranted to explore this possibility.
Finally, we would like to comment on some limitations of our study. First, we have to point out that serum samples for bile acid analysis were available for only 56 of the 150 initial patients. Given the subanalysis nature of the current study, it would be ideal to maintain the same number of patients used in the initial study. Unfortunately, administrative issues, as outlined earlier, did not allow this to happen. Lastly, we would like to cite the significant difference in the percentage of patients having concomitant inflammatory bowel disease between the group of patients that were analyzed for bile acid composition and the group of patients that were not (93% versus 68%). In theory, inflammation in the colon can influence the intestinal absorption and bacterial degradation of UDCA. However, the currently available data show that no significant differences with respect to the bile acid composition were detected between the UDCA-treated PSC patients who had colitis and those who did not.27 The small number of patients without inflammatory bowel disease included in our study (n = 4) did not allow us to perform a comprehensive statistical analysis in order to check this hypothesis.
In summary, we suggest that high-dose UDCA treatment results, as expected, in total bile acid expansion and significant UDCA enrichment in patients with PSC. LCA is markedly increased as well. Further studies are needed in order to fully understand UDCA-induced liver damage in patients with PSC.
- 4High-dose ursodeoxycholic acid as a therapy for patients with primary sclerosing cholangitis. Am J Gastroenterol 2001; 96: 1558-1562., , , , .Direct Link:
- 21Relationship between biliary and serum bile acids and response to ursodeoxycholic acid in patients with primary biliary cirrhosis. Am J Gastroenterol 1998; 93: 1498-1504., , , , , , et al.Direct Link: