Effect of high-dose ursodeoxycholic acid on its biliary enrichment in primary sclerosing cholangitis



Ursodeoxycholic acid (UDCA) has beneficial effects in cholestatic liver diseases. In primary sclerosing cholangitis (PSC), there is evidence that high doses (±20 mg/kg) of UDCA may be more effective than average doses. Biliary enrichment of UDCA at such high doses may represent the decisive factor for its beneficial effect. Up to now it is not clear how high-dose UDCA correlates with its biliary enrichment and whether bacterial degradation of large amounts of UDCA may lead to an increased bacterial formation of more toxic hydrophobic bile acids. We determined the biliary bile acid composition in 56 patients with PSC including 30 patients with repeat bile samples treated with various doses of UDCA. At a UDCA dose of 10–13 mg/kg/d (n = 18) biliary UDCA represented 43.1% + 0.3% (mean + SD) of total bile acids; at a UDCA dose of 14–17 mg/kg (n = 14), its biliary content increased to 46.9% + 0.3%, at 18–21 mg/kg (n = 34) to 55.9% + 0.2%, at 22–25 mg/kg (n = 12) to 58.6% + 2.3%, and at 26–32 mg/kg (n = 8) to 57.7% + 0.4%. During UDCA treatment, the biliary content of all other bile acids was unchanged or decreased. In conclusion, biliary enrichment of UDCA increases with increasing dose and reaches a plateau at 22–25 mg/kg. There was no increase of toxic hydrophobic bile acids. If biliary enrichment of UDCA represents the decisive factor for its clinical effect, it seems likely that UDCA doses of up to 22–25 mg/kg may be more effective than lower doses. (HEPATOLOGY 2004;40:693–698.)

Ursodeoxycholic acid (UDCA) has beneficial effects in cholestatic liver diseases. In primary biliary cirrhosis, UDCA treatment slows progression of the disease and improves survival.1–6 In primary sclerosing cholangitis (PSC), UDCA improves laboratory parameters,7–12 and its effect on the outcome is currently being evaluated. There is evidence11, 12 that high doses (≥20 mg/kg/d) of UDCA may be more effective than average doses (15 mg/kg/d). The biliary enrichment of UDCA, which may represent the key factor for its beneficial effect,13 has not been studied at such high doses. The intestinal absorption of UDCA is incomplete14, 15 and the bacterial degradation of unabsorbed UDCA in the intestine may affect its own biliary enrichment and that of other bile acids.16, 17 The aim of the present study was to determine the optimal dose of UDCA with respect to its biliary enrichment.


UDCA, ursodeoxycholic acid; PSC, primary sclerosing cholangitis; ERC, endoscopic retrograde cholangiography.

Patients and Methods

The selection criteria for enrolling patients with PSC in the study included typical endoscopic retrograde cholangiographic findings, serum alkaline phosphatase activity of at least twice the normal range, negative antimitochondrial antibody, and a liver biopsy compatible with the diagnosis of PSC. Criteria for exclusion were decompensation of cirrhosis, patients in whom liver transplantation was foreseen, patients with a history of neoplastic disease and/or coexisting hepatic disease. Table 1 shows the clinical data of the patients.

Table 1. Patient Data at Entry Into Study (n = 56)
Age, yr40.2 ± 1.7
Weight, kg74.2 ± 1.4
Ulcerative colitis, n (%)29 (52%)
Crohn's disease, n (%)2 (4%)
Alkaline phosphatase, IU348 ± 32.1
Gamma-glutamyltransferase, IU75.0 ± 12.0
Alanine aminotransferase, IU43.3 ± 4.6
Serum bilirubin, mg/dL2.0 ± 0.4
Serum albumin, g/L39.0 ± 1.0
International normalized ratio1.03 ± 0.02

Of 56 patients, 29 had ulcerative colitis, and 2 had Crohn's disease. Patients with ulcerative colitis and previous colectomy with ileoanal pouch were excluded. Twenty-six patients with ulcerative colitis and 1 with Crohn's disease received 5-aminosalicylic acid (3 g/d), and 7 with ulcerative colitis and the 2 with Crohn's disease received intermittently corticosteroids (methylprednisolone, 10-50 mg/d) for periods of up to 3 months. No differences between patients treated with 5-aminosalicylic acid and/or corticosteroids and those who needed no treatment of their inflammatory bowel disease were noted. The biliary system was visualized by endoscopic retrograde cholangiography (ERC) in each patient. In all patients with narrowing of the common bile duct, ERC was repeated at yearly intervals.19 In patients without narrowing of the common duct at the first visit, ERC was repeated whenever alkaline phosphatase and/or gamma-glutamyltransferase and/or serum aminotransferases and/or serum bilirubin increased by more than 20%. For prevention of bacterial cholangitis as a consequence of ERC, antibiotics were administered in each of the patients.

All patients received 3 × 2 g of mezlocillin (Baypen) intravenously, both on the day before and after the ERC. In addition, 50 mg of netilmycin (Certomycin) were added to each 20-mL portion of the contrast medium iopamidol (Solutrast). In patients with a total or subtotal stenosis of a major duct (stenosis of common duct and/or of the hepatic ducts within 2 cm of the bifurcation) and biochemical evidence of cholestasis, a dilatation of the stenoses was performed. Balloon dilatation was repeated at 4-week intervals up to success, as assessed by opening of the stenosis at repeat cholangiography. Details of the treatment regimen have previously been published.19 When the original stenosis was open at ERC, the next control ERC was performed after 3 months and afterward after doubling of the time interval (6 and 12 months). In general, patients with dominant stenosis needed 1 repeat ERC with dilatation per year to prevent stenosis. During ERC, the bile duct was selectively intubated, and bile was collected via the intubation catheter and immediately frozen at −20°C. Since cholestasis may lead to reduced absorption of UDCA20, and increasing evidence suggests that higher doses may be more effective,11, 12 in our patients the UDCA dose has been increased to 18–21 mg/kg per day.12 To study the effect of even higher UDCA doses,11 in 20 arbitrarily selected patients the dose has been increased to 22–32 mg/kg (in three divided doses). The 22 patients without dominant stenosis in whom one single bile sample was obtained belong to the group of patients in whom the ERC was performed due to an increase of alkaline phosphatase, gamma-glutamyltransferase , or serum bilirubin, and who did not have a dominant stenosis. The 30 patients with repeat bile samples belong to the group of patients with dominant stenoses who needed repeat ERC for therapeutic purposes. In patients with biliary obstruction, bile samples were not collected before at least 1 week after relief of the obstruction. In 11 out of 30 patients, baseline bile samples before treatment with UDCA were obtained. In 12 out of 30 patients, repeat bile samples were obtained at 4-week intervals at the same UDCA dose (8 with 2 bile samples, and 4 with 4 bile samples, each). In all 30 patients the UDCA dose was increased to 18–32 mg/kg, and repeat bile samples were obtained in 3- to 12-month intervals. In 13 of these patients, the dose was increased from 10–13 mg/kg to 18–22 mg/kg, in 7 patients from 13–17 mg/kg to 18–22 mg/kg , in 7 patients from 18–22 mg/kg to 25–30 mg/kg, and in another 3 patients from 23–27 to 28–32 mg/kg.

Analytical Methods

The methods for analysis of biliary bile acids have been described earlier.21, 22 After solvolysis of bile acid sulfates and hydrolysis of bile acid glucuronides by glucuronidase, the bile acids were separated on DEAP-LH 20 into nonamidated bile acids, and glycine and taurine conjugates. Gas-liquid chromatography of methylated and trimethylsilyl-substituted bile acid derivatives was performed on 4.5 m 0.5% HiEff 8BP (Applied Science Laboratories Inc., State College, PA) conventional glass columns and in addition on 50 m OV 101 (Machery and Nagel, Dueren, Germany) glass capillary columns. An HP 5710A gas chromatograph (Hewlett Packard, Bad Homburg, Germany) equipped with a flame ionization detector was used.

Statistical Analysis.

Continuous data were compared with the nonparametric Wilcoxon rank-sum test. Results are given as mean + SD.

Informed consent was obtained from each patient, and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki, as reflected in a priori approval by the institution's human research review committee.


Intraindividual variation of UDCA enrichment was studied in 12 patients (Fig. 1). In 8 patients with 2 bile samples obtained at an average UDCA dose of 20.11 (range, 16.0–25.0) mg/kg/d, intraindividual UDCA enrichment varied up to 19.84% (mean, 11.91%; range, 4.14%–19.84%). In 4 patients with 4 values at an average UDCA dose of 22.64 (range, 20.84–24.70) mg/kg, maximal intraindividual variation of UDCA in bile was 2.07%, 5.70%, 16.46%, and 18.19% (Fig. 1). On the basis of the dose, 5 dose groups were evaluated (Table 2): 10–13 mg/kg, 14–17 mg/kg, 18–21 mg/kg, 22–25 mg/kg, and 26–32 mg/kg. With increasing dose, UDCA enrichment in bile increased and reached a plateau at 22–25 mg/kg (Fig. 2). An almost identical curve was obtained when all the lower values of patients with multiple samples were excluded. At a UDCA dose of 10–13 mg/kg (n = 18) biliary UDCA ranged from 20.2% to 62.0% of total bile acids, at 14–17 mg/kg (n = 14) from to 26.3% to 66.8%, at 18–21 mg/kg (n = 34) from 31.5% to 68.8%, at 22–25 mg/kg (n = 12) from 42.0% to 71.9%, and at 26–32 mg/kg (n = 8) from 42.5% to 73.4%. No differences were observed between patients with and without colitis. In the 6 study groups, serum parameters of cholestasis (alkaline phosphatase, gamma-glutamyltransferase, bilirubin) were not significantly different. Up to a UDCA dose of 22 mg/kg, in all patients with repeat bile samples a further increase of the dose led to an increased biliary enrichment of UDCA, whereas this was not the case at doses >22 mg/kg (Fig. 3). With increasing UDCA dose, the sum of all other bile acids in bile decreased with the same percentage in which UDCA increased (Table 2). The individual bile acids cholic acid and chenodeoxycholic acid and their bacterial degradation products deoxycholic acid and lithocholic acid decreased slightly (not significant; Table 2). With increasing dose, the conjugation of UDCA and all other bile acids with glycine changed little (not significant; Table 3).

Figure 1.

Variation of UDCA enrichment in individuals treated with a given dose of UDCA. In 8 patients, 2 bile samples and in 4 patients 4 bile samples were obtained while the UDCA dose remained unchanged. At a given dose, UDCA enrichment in bile samples obtained at different time points in 1 patient varied up to 20%. UDCA, ursodeoxycholic acid.

Table 2. Biliary Bile Acid Composition
Bile AcidBefore Treatment (n = 11)UDCA Treatment mg/kg/d
10–13 (n = 18)14–17 (n = 14)18–21 (n = 34)22–25 (n = 12)26–32 (n = 8)
  • NOTE. Values are means ± SD.

  • Abbreviations: CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, lithocholic acid.

  • *

    Significantly different from values before treatment (P < .05).

  • Significantly different from the value at lower dose.

CA, mmol/L (%)1.21 ± 0.100.88 ± 0.010.49 ± 0.010.67 ± 0.010.82 ± 0.120.53 ± 0.08
 (64.1 ± 5.1)(24.1 ± 0.3)*(27.0 ± 0.3)*(18.8 ± 0.2)*(16.7 ± 2.5)*(21.8 ± 3.4)*
CDCA, mmol/L (%)0.53 ± 0.070.79 ± 0.040.33 ± 0.030.65 ± 0.040.92 ± 0.080.43 ± 0.05
 (28.1 ± 3.8)(21.7 ± 1.2)*(18.2 ± 1.7)*(18.2 ± 1.0)*(18.8 ± 1.7)*(17.5 ± 2.2)*
DCA, mmol/L (%)0.12 ± 0.050.37 ± 0.070.13 ± 0.030.21 ± 0.030.25 ± 0.010.06 ± 0.03
 (6.4 ± 2.8)(10.1 ± 1.9)(7.0 ± 1.4)(6.0 ± 0.9)(5.2 ± 0.1)(2.6 ± 1.2)
UDCA, mmol/L (%)0.02 ± 0.011.56 ± 0.010.86 ± 0.011.99 ± 0.012.88 ± 0.111.41 ± 0.01
 (1.0 ± 0.7)(43.1 ± 0.3)*(46.9 ± 0.3)*(56.0 ± 0.2)*(58.6 ± 2.3)*(57.7 ± 0.4)*
LCA, mmol/L (%)0.01 ± 0.010.04 ± 0.010.02 ± 0.010.03 ± 0.010.04 ± 0.020.01 ± 0.01
 (0.3 ± 0.1)(1.0 ± 0.3)(0.9 ± 0.3)(1.0 ± 0.2)(0.8 ± 0.4)(0.4 ± 0.3)
Figure 2.

Biliary enrichment of UDCA at various UDCA doses. The data of 86 bile samples obtained in 54 patients were included. The enrichment of UDCA in bile reaches a plateau at 22–25 mg/kg/d UDCA. UDCA, ursodeoxycholic acid.

Figure 3.

Biliary enrichment of UDCA in patients who were treated with 2 different UDCA doses. In all patients treated with doses <22 mg/kg/d, an increase of the dose led to an increased biliary enrichment of UDCA. UDCA, ursodeoxycholic acid.

Table 3. Conjugation Profile of Biliary Bile Acids (%)
Bile AcidConjugateBefore Treatment (n = 11)UDCA Treatment mg/kg/d
10–13 (n = 18)14–17 (n = 14)18–21 (n = 34)22–25 (n = 12)26–32 (n = 8)
  1. NOTE. Values are means ± SD. During treatment, glycine conjugation of all bile acids was significantly increased (P < .05).

  2. Abbreviations: CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, lithocholic acid.

CAGlycine52.4 ± 6.982.6 ± 8.972.8 ± 2.881.9 ± 4.478.1 ± 2.377.1 ± 1.6
 Taurine46.0 ± 7.116.8 ± 9.125.8 ± 3.616.8 ± 4.020.4 ± 2.922.6 ± 1.6
CDCAGlycine47.7 ± 6.684.6 ± 7.172.9 ± 2.984.1 ± 5.178.3 ± 3.880.6 ± 1.6
 Taurine51.8 ± 6.515.4 ± 7.125.4 ± 3.116.0 ± 5.121.2 ± 3.919.3 ± 1.7
UDCAGlycine68.6 ± 11.488.0 ± 4.184.6 ± 1.190.7 ± 2.788.1 ± 1.089.1 ± 0.7
 Taurine31.4 ± 11.412.0 ± 4.114.8 ± 1.49.3 ± 2.711.9 ± 1.010.8 ± 0.7
DCAGlycine55.4 ± 5.873.2 ± 7.564.8 ± 5.672.4 ± 6.768.5 ± 4.774.0 ± 2.6
 Taurine43.2 ± 6.426.8 ± 7.530.3 ± 4.327.6 ± 6.731.5 ± 4.725.3 ± 2.8
LCAGlycine53.466.659.0 ± 9.889.4 ± ± 8.1
 Taurine46.633.330.4 ± 5.510.6 ± ± 6.4
TotalGlycine51.5 ± 7.084.4 ± 7.078.2 ± 2.187.9 ± 3.584.8 ± 1.884.8 ± 1.1
 Taurine47.3 ± 7.115.5 ± 7.120.6 ± 5.811.9 ± 3.515.1 ± 1.915.1 ± 1.1


UDCA has been shown to improve liver function tests, liver histology, and survival in primary biliary cirrhosis,1–6 and similar effects have been observed in other cholestatic diseases.7–12, 23–25 The mechanism of action of UDCA in hepatobiliary diseases is still not completely understood.26–38 Decreased intestinal absorption of hydrophobic, hepatotoxic bile acids,39, 40 and biliary enrichment with hydrophilic, nontoxic UDCA changes the balance of biliary bile acids in favor of nontoxic hydrophilic bile acids, and this change may represent one of the mechanisms of action of UDCA.13 In patients, studies on biliary enrichment are difficult to perform. Ideally, biliary enrichment of UDCA should be studied at regular intervals coinciding with scheduled dosage changes. To do such a study, however, obviously will be clinically difficult. All bile samples have been collected after selective intubation of the bile duct during clinically indicated endoscopic measures. After oral administration of UDCA, there is a considerable interindividual variation in the biliary enrichment of this bile acid, which makes it difficult to determine the optimal dose of UDCA in the individual patient. At low UDCA doses, its biliary enrichment ranges from very low values to values that in other patients are achieved only at high-dose treatment. In the low-dose group (10-13 mg/kg), the difference between the highest and lowest biliary enrichment rate of UDCA was 40%, and with increasing dose this value decreased (Fig. 2), indicating that the biliary enrichment is especially unpredictable at low UDCA doses. The minimum percentage of UDCA in bile needed for a therapeutic effect is unknown. It is obvious that if UDCA enrichment represents the key for its clinical effect, in some patients a low dose may be sufficient, and in others it may be insufficient. Our data indicate that by increasing the UDCA dose, the number of patients with presumably sufficient biliary UDCA enrichment increases.

The wide range of biliary enrichment of UDCA makes it difficult to determine the dose above which no further benefit in enrichment can be expected. The analysis of 86 bile samples obtained with different UDCA doses indicates that at 22–25 mg/kg UDCA, the biliary enrichment curve reaches a plateau. This means that above this dose the majority of patients will not respond to a further increase of the dose with a further increase of UDCA enrichment in bile (Fig. 2). However, more data at doses above 26 mg/kg are needed to confirm this assumption. The extent to which biliary enrichment with UDCA varies from day to day in a given patient up to now was unknown.41 In the present study the variation of enrichment at a given dose in the same patient was approximately half of that in different subjects (Figs. 1 and 2). Therefore, paired data obtained in the same patient allow more precise evaluation of the effect of an increased dose. In general, patients with low UDCA percentage at a low dose also had a relatively low percentage at higher doses (Fig. 3). A UDCA dose of up to 22 mg/kg in all 27 individuals tested a further increase of the dose was followed by an increase of UDCA in bile (Fig. 3). In 3 patients in whom UDCA doses of >22 mg/kg were further increased this increase in dose led to in part to an increase and in part to a decrease of UDCA. These data obtained in patients with paired bile samples are in agreement with those of the whole study group (Fig. 2) and support the concept that for the majority of patients the optimal dose is between 22 and 25 mg/kg. This finding does not exclude the possibility that in individual patients even higher doses of UDCA may have an additional effect on its biliary enrichment. In a recent study in patients with primary biliary cirrhosis, after 750 mg/d tauroursodeoxycholic acid biliary enrichment was 32%, and after 750 mg/d UDCA it was 29%.42 The present data in PSC patients indicate that by increasing the UDCA dose, much higher biliary enrichment values for UDCA can be obtained. Absorption of orally administered UDCA is slow and incomplete.14, 15 Therefore it can be expected that at high doses of UDCA relatively large amounts of unabsorbed UDCA are subject to bacterial degradation. This may lead to an increased bacterial formation of 7ketolithocholic acid and lithocholic acid and by the hydroxylation of 7ketolithocholic acid in the liver, also to an increase of chenodeoxycholic acid.17, 18, 43, 44 In the patients studied, biliary chenodeoxycholic acid and lithocholic acid decreased. In agreement with previous recommendations,45 all our patients received antibiotics during endoscopic procedures and therefore the bacterial formation of secondary bile acids may be somewhat reduced. Since bile acids are transported into the bile by specific transporters not used by antibiotics, a direct effect of antibiotics on bile composition seems unlikely. The data obtained in this study indicate that under the given conditions there is no increase of other bile acids, even at very high UDCA doses (Table 2). When administered orally, unconjugated UDCA on the first pass through the liver is rapidly conjugated with glycine or taurine.46 After UDCA, bile was mainly enriched with the glycine conjugate of UDCA, and at very high UDCA doses this percentage increased only a little (Table 3). Several studies have suggested that the taurine conjugates may have more pronounced hepatoprotective effects than UDCA.28, 29, 47 Studies with glycine-conjugated UDCA, however, indicate a similar hepatoprotective effect for taurine and glycine conjugates of UDCA.48 The little differences in conjugation between the average and high doses of UDCA make it unlikely that this factor is of relevance when patients with PSC are treated with high doses of UDCA. After high-dose UDCA treatment of patients with PSC, its biliary content increased on average to approximately 60%, which is much more than what may be achieved with conventional doses. These findings may represent a rationale for the improved efficacy of such high UDCA doses.