A Randomized, Dose-Finding, Proof-of-Concept Study of Berberine Ursodeoxycholate in Patients With Primary Sclerosing Cholangitis : Official journal of the American College of Gastroenterology | ACG

Secondary Logo

Journal Logo


A Randomized, Dose-Finding, Proof-of-Concept Study of Berberine Ursodeoxycholate in Patients With Primary Sclerosing Cholangitis

Kowdley, Kris V. MD1; Forman, Lisa MD2; Eksteen, Bertus PhD, MBChB3; Gunn, Nadege MD4; Sundaram, Vinay MD5; Landis, Charles MD, PhD6; Harrison, Stephen A. MD7; Levy, Cynthia MD8; Liberman, Alexander PhD9; Di Bisceglie, Adrian M. MD9; Hirschfield, Gideon M. PhD, MB BChir10

Author Information
The American Journal of Gastroenterology: November 2022 - Volume 117 - Issue 11 - p 1805-1815
doi: 10.14309/ajg.0000000000001956



Primary sclerosing cholangitis (PSC) is a rare inflammatory, fibrotic disease involving intrahepatic and extrahepatic bile ducts (1). Patients with PSC are at risk of progression to biliary cirrhosis, hepatic decompensation, and cholangiocarcinoma (2). PSC is very often associated with inflammatory bowel disease (IBD), including but not limited to ulcerative colitis or Crohn disease (1). Although the pathogenesis of PSC is uncertain, there is evidence of an inciting role for bacteria or bacterial products originating from the gut and a subsequent autoinflammatory response with some genetic determinants (3,4).

There is no known effective therapy for PSC and its progression seems to be unaffected by immunosuppressive therapies, including those used to treat underlying IBD (1). Ursodeoxycholic acid (UDCA) has been extensively tested in PSC and is associated with reductions in serum levels of alkaline phosphatase (ALP), a key biomarker of cholestasis and bile duct injury, but has not been shown to slow the progression of liver or biliary disease associated with PSC (5,6). Furthermore, a trial of high-dose UDCA was harmful, with increased liver-related adverse outcomes in patients receiving 28–30 mg/kg/d of UDCA compared with placebo recipients (7). As a result, there is continued controversy regarding the use of UDCA in PSC.

Berberine ursodeoxycholate (HTD1801) is an ionic salt of berberine (BBR) and UDCA and represents a new molecular entity offering the possibility of combination therapy for PSC—thus, it is more than just a physical mixture of these 2 agents (8). The molecular weights of the BBR cation and UDCA anion in HTD1801 are approximately 336 and 391 g/mol, respectively. The mechanisms of action of UDCA have been well studied in a variety of cholestatic liver diseases, including PSC (9). In addition to the known choleretic, cytoprotective, and anti-inflammatory effects of UDCA, BBR has substantial antimicrobial activity and is an anti-inflammatory agent (10). In particular, BBR has activity against Klebsiella pneumoniae, a bacterium believed to be a pathobiont that is causally associated with the development of PSC (4). Furthermore, BBR is excreted in part unchanged in bile, which may be beneficial in PSC, because bacteria and yeast are often found in the bile of patients with PSC even when they do not have clinical evidence of ascending cholangitis (11). In a MDR2−/− mouse model of cholestatic disease, treatment with BBR improved cholestatic liver injury with reductions in transaminases and ALP (12). A pilot study investigating the effects of berberis vulgaris oxymel in patients with PBC (n = 12) or PSC (n = 13) found that daily doses at 0.5 mL/kg in addition to underlying therapy resulted in significant reductions in ALP, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and gamma glutamyltransferase (GGT) after 3 months of treatment (13). When administered in the form of the ionic salt, both BBR and UDCA have increased solubility and bioavailability. This proof-of-concept study of HTD1801 examined the safety and tolerability of HTD1801 as well as change in serum ALP levels in patients with PSC with and without IBD.


A phase 2, randomized, placebo-controlled trial was conducted evaluating 2 doses of HTD1801 (500 mg BID and 1,000 mg BID) versus placebo (ClinicalTrials.gov identifier: NCT03333928). A 1000-mg dose of HTD1801 delivers approximately 538 mg of UDCA anion and 462 mg of BBR cation. The total duration of this study was 18 weeks divided into three 6-week periods, the first being a placebo-controlled period, with the primary end point being change from baseline in ALP at the end of those 6 weeks. This was followed by a treatment extension period wherein all patients receiving HTD1801 maintained their dose while patients originally randomized to placebo were rerandomized to receive either HTD1801 500 or 1,000 mg BID. This was followed by a randomized treatment withdrawal period for the final 6 weeks wherein patients on treatment with HTD1801 were rerandomized (1:1) to placebo or their prior dose of the study drug for a further 6 weeks (Figure 1).

Figure 1.:
Consort diagram. Fifty-nine patients were randomized, with 4 patients withdrawing before receiving treatment with the remaining patients constituting the safety population (N = 55). One patient who had been randomized to the HTD1801 1,000 mg BID group withdrew consent before any efficacy evaluations and was therefore excluded from the mITT population (n = 54). Each period was 6 weeks in duration: period 1: day 0-week 6; period 2: week 6-week 12; and period 3: week 12-week 18. The study primary end point, change from baseline in ALP, was evaluated at the end of period 1 (week 6). During period 2, HTD1801 randomized patients maintained their period 1 dose while placebo patients were evenly randomized to either HTD1801 dose. During period 3, patients were rerandomized to maintain their period 2 dose or placebo. ALP, alkaline phosphatase; BID, twice daily; HTD1801, berberine ursodeoxycholate; mITT, modified intent-to-treat.

Patients with a known diagnosis of PSC based on magnetic resonance cholangiopancreatography (MRCP) or endoscopic retrograde cholangiopancreatography (ERCP) findings were eligible for inclusion in this study. To be eligible for randomization into this study, serum ALP levels at screening had to be elevated at least more than 1.5 times the upper limit of normal (ULN). Patients were excluded if they had a known or suspected dominant biliary stricture or any other identifiable cause of liver disease, including autoimmune hepatitis, small-duct PSC, IgG4-related cholangitis, or evidence of decompensated liver disease. Patients with severely active IBD or a flare in colitis activity within 90 days of screening were ineligible from study participation. Medications known to be strong inhibitors of P-glycoprotein, cytochrome P450 3A4, cytochrome P450 2D6, and P-glycoprotein substrates with a narrow therapeutic index were prohibited during the course of this study due to unknown potential interactions with BBR at the time of study conduct.

Initially, patients who were taking UDCA were required to discontinue its use at least 6 weeks before entering the study. However, to facilitate study enrollment, the protocol was subsequently amended to allow those taking UDCA to continue until the day 0 visit, when it had to be stopped. Prior UDCA usage was defined as any use of UDCA reported in medical history before the start of the study. Laboratory tests were performed every 2 weeks on therapy, including serum chemistry and blood counts, and for those patients with a history of IBD, the status of the IBD was checked using the partial Mayo Score (14), serum levels of C-reactive protein, and fecal calprotectin levels.

Patients were recruited from 27 sites in the United States and Canada. The first patient was randomized in June 2018 while the last patient completed their participation in this study in August 2020. This study protocol was approved by local and regional ethical review boards and regulatory agencies and was conducted in accordance with the Good Clinical Practice Guidelines.

Statistical methods

The sample size for this study was originally planned to be 90 patients. In view of the protracted period necessary to enroll patients into this study and challenges related to the COVID-19 pandemic, the target sample size was lowered to 63. A sample size of 21 patients per group was estimated to provide 80% power to detect a mean difference of 180 U/L for each dose group vs placebo, with an assumed SD of change from baseline in ALP of 200 U/L and alpha of 0.05. However, the achieved sample size was slightly less than the recruitment target (55 vs 63 patients). Nevertheless, statistical significance was achieved in part because of an observed SD that was lower than the assumed SD.

This study included 3 randomizations, 1 for each period. In period 1, patients were randomized to placebo BID, HTD1801 500 mg BID, or HTD1801 1,000 mg BID. Randomization was stratified according to a coexisting diagnosis of IBD and UDCA status. The UDCA stratification factor was added part way through this study after 42 patients had been randomized to account for the protocol amendment, which no longer required a 6-week UDCA washout. UDCA status was retroactively applied as a stratification factor to those 42 patients who had already been randomized in period 1. This process of adding in the UDCA stratification resulted in 24 of 59 patients being randomized to HTD1801 1,000 mg BID while 15 of 59 and 16 of 59 were randomized to HTD1801 500 mg BID and placebo, respectively.

In period 2, patients receiving placebo were rerandomized to HTD1801 500 mg BID or 1,000 mg BID while those already assigned to HTD1801 maintained their period 1 dose. After the completion of period 2, patients were rerandomized to either continue their period 2 dose (i.e. HTD1801 to HTD1801) or to cross over to placebo (i.e. HTD1801 to placebo). This resulted in approximately half of the patients remaining on active treatment at their previous dose and half receiving placebo and withdrawing from HTD1801. Neither UDCA nor IBD status was a factor involved in these subsequent randomizations because the primary end point was evaluated at the end of period 1. All HTD1801 and placebo were provided as matching white tablets.

Efficacy analyses were based on the modified intent-to-treat (mITT) population, and safety analyses were based on the safety population. The mITT population consisted of all randomized patients who received at least 1 dose of the study drug and who had at least 1 postdose efficacy assessment. Patients in this population were analyzed according to the treatment group to which they were randomized. The safety population consisted of all patients who received at least 1 dose of study treatment. Patients in this population were analyzed according to the treatment they received at the time of the event being summarized. That is, for patients who received different treatments in different study periods, the safety events are summarized based on the treatment received at the time of the event.

Descriptive statistics were used to summarize the data from this study, with hypothesis testing performed in period 1 for the ALP and GGT end points. Testing was based on a mixed model with repeated measures looking at absolute change from baseline as the response variable and fixed effects for treatment, IBD diagnosis, UDCA status, visit, and treatment by visit interaction, with the associated baseline laboratory parameter as a covariate. An unstructured covariance matrix was used along with the Kenward-Roger method. Prespecified subgroup analyses included analyses based on baseline IBD status and history of prior UDCA usage. Analyses were performed using SAS system version 9.4 or higher.


Eighty-eight patients were screened, 59 were randomized and 55 received at least 1 dose of the study drug (safety population) as shown in Figure 1. The mITT population constituted of 54 patients who had at least 1 ALP measurement during treatment because 1 patient who had been randomized to the HTD1801 1,000 mg BID group withdrew consent before any efficacy evaluations. The baseline characteristics of the study population are shown in Table 1. Briefly, their mean age was 43 years and most of them were male (n = 32, 58%) and White (n = 46, 84%). Thirty-five patients (64%) had a history of inflammatory bowel disease and 22 (40%) had a history of UDCA use. Of the 22 patients with a history of UDCA use, 3 discontinued UDCA at least 32 days before study entry, 17 stopped at study entry, and 2 patients inadvertently continued through study start ceasing use at day 16 and day 43 (both having been randomized to placebo). At the time of enrollment, the mean serum ALP for the safety population was 375 U/L, ranging between 122 and 1048 U/L. Four patients (6.8%) had screening ALP >1.5x ULN but day 0 ALP <1.5x ULN (placebo, n = 2; HTD1801 1,000 mg BID, n = 2).

Table 1.:
Baseline demographic characteristics by initial treatment group (safety population)

Period 1 (day 0 to week 6): placebo-controlled period

Serum ALP promptly decreased with active therapy so that after 2 weeks, there was a 21% mean decrease in the HTD1801 1,000 mg BID group and a 17% decrease in the HTD1801 500 mg BID group in contrast to a 25% increase in the placebo group. These changes persisted throughout the first 6-week treatment period. The primary end point of this study was achieved at week 6 with a least squares mean change in ALP from baseline of −37 (SE, 36.2) U/L (P = 0.019 vs placebo), −53 (43.7) U/L (P = 0.016 vs placebo), and 98 (42.5) U/L for the HTD1801 1,000 mg BID, HTD1801 500 mg BID, and placebo groups, respectively. The mean (SE) change from baseline at week 6 was −73 (29.8) U/L, −71 (35.3) U/L, and 94 (67.6) U/L for the HTD1801 1,000 mg BID, HTD1801 500 mg BID, and placebo groups, respectively. The absolute values and change from baseline in ALP are shown in Figures 2a,b. Furthermore, a greater percentage of patients in the HTD1801 treatment groups achieved clinically relevant ALP reduction thresholds (ALP <1.5x ULN, at least 50% reduction in ALP, or normalization of ALP) than placebo as shown in Figure 2c. A sensitivity analysis was performed post hoc (see Supplemental Figure 1, Supplementary Digital Content 1, https://links.lww.com/AJG/C650), excluding patients with day 0 ALP <1.5x ULN, which shows consistent results with Figure 2C.

Figure 2.:
Changes in ALP over 6 weeks of treatment. (a) ALP over time during the placebo-controlled phase of the study (period 1). (b) Change from baseline in ALP during the placebo-controlled phase of the study (period 1). *P < 0.05. P-value derived from a MMRM testing absolute change from baseline, with fixed effects for treatment, IBD diagnosis, UDCA status, visit, and treatment-by-visit interaction, with the baseline ALP value as a covariate. (c) Percentage of patients achieving relevant ALP thresholds during the placebo-controlled phase of the study (period 1). ALP, alkaline phosphatase; BID, twice daily; IBD, inflammatory bowel disease; MMRM, mixed model repeated measures; UDCA, ursodeoxycholic acid; ULN, upper limit of normal.

Reductions in ALP with active treatment closely paralleled reductions in GGT, AST, and ALT. As shown in Figure 3, GGT, AST, and ALT also improved after 6 weeks of HTD1801 treatment in contrast to each of these 3 markers worsening in placebo-treated patients.

Figure 3.:
Change from baseline to week 6 in GGT, AST, and ALT. Change from baseline in GGT (a), AST (b), and ALT (c) at the end of the placebo-controlled phase of the study (period 1). ALT, alanine aminotransferase; AST, aspartate aminotransferase; BID, twice daily; GGT, gamma glutamyltransferase.

To distinguish the beneficial treatment effect of HTD1801 from the effect of withdrawal from UDCA during period 1, an analysis was conducted based on prior UDCA usage, as defined in the methods section. As shown in Figure 4, patients randomized to placebo who had prior UDCA usage on average had worsening ALP in contrast to patients randomized to HTD1801 whose ALP remained near baseline levels. Note that 17 of 22 patients in the prior UDCA group had discontinued their UDCA immediately before receiving the study drug, so for those randomized to placebo, this amounted to withdrawal of their UDCA. In patients with no prior UDCA usage, patients randomized to placebo experienced little change on average in ALP from baseline, whereas both HTD1801 groups had reductions in ALP sustained throughout period 1. Supplemental Figure 2, Supplementary Digital Content 1, https://links.lww.com/AJG/C650 displays waterfall plots of individual subject change from baseline to week 6 in ALP by treatment group and by prior UDCA usage. As can be seen across treatment groups, in patients with no prior UDCA usage, ALP was typically reduced from baseline in contrast to patients with prior UDCA usage. In addition, with HTD1801 1,000 mg, in patients with no prior UDCA usage, greater reductions in ALP were observed with no increases in ALP in any patients.

Figure 4.:
Changes in ALP over 6 weeks by prior UDCA usage. Prior UDCA usage was defined as any UDCA usage before the start of the study determined by a review of the medical history. Forty percent of patients had a history of prior UDCA usage. (a) ALP throughout period 1 for patients with a history of prior usage of UDCA. (b) ALP throughout period 1 for patients with no history of UDCA usage. ALP, alkaline phosphatase; BID, twice daily; UDCA, ursodeoxycholic acid.

To evaluate possible effects of IBD on treatment, the median percentage of decrease in ALP was compared in patients with and without IBD. Only 19 of 55 patients (35%) did not have IBD, making the number in each treatment group quite small. Nonetheless, the median decrease in ALP by week 6 was similar in both active treatment groups in those who did and did not have IBD (−22% and −22% with HTD1801 500 mg BID and HTD1801 1,000 mg BID, respectively, compared with −10% and −24%). In patients receiving placebo, median ALP increased by 34% at week 6 in those with IBD and decreased slightly in the 5 patients (−8%) who did not have IBD.

Period 2 (week 6 to week 12): treatment extension period

In period 2, patients initially randomized to placebo were rerandomized to receive either HTD1801 500 mg BID or HTD1801 1,000 mg BID while those already on HTD1801 continued their prior doses. As shown in Figure 5a, rerandomized placebo patients experienced decreases in ALP after initiation of HTD1801 (mean [SE] change from week 6 to week 12, placebo to HTD1801 500 mg BID: −34 [68] U/L, placebo to HTD1801 1,000 mg BID: −189 [102] U/L). Patients who continued to receive HTD1801 maintained their prior reductions in ALP, as shown in Figure 5b, demonstrating ongoing durability of response with HTD1801.

Figure 5.:
Changes in ALP during the treatment extension (period 2) and randomized withdrawal (period 3) phases. (a) Placebo randomized patients transitioning to HTD1801 during period 2. (b) HTD1801 randomized patients maintaining their period 1 dose through period 2. (c) Randomized withdrawal of HTD1801 to placebo in period 3. (d) HTD1801 randomized patients maintaining their period 2 dose through period 3. ALP, alkaline phosphatase; BID, twice daily; HTD1801, berberine ursodeoxycholate.

Period 3 (week 12 to week 18): randomized withdrawal period

After the completion of period 2, all patients were rerandomized to receive either placebo or their prior dose of HTD1801. As shown in Figure 5c, patients in whom HTD1801 was withdrawn experienced a rise in ALP as early as week 14 (i.e. 2 weeks after stopping HTD1801), which was sustained to the end of period 3. Patients who were maintained on HTD1801 either continued to experience a further gradual decrease in ALP or sustained their ALP response achieved during period 2 further indicative of a durable effect with HTD1801.


Owing to the extension and randomized cross-over nature of the study design, it was not always clear whether reported adverse events were associated with the current treatment or a prior treatment. For consistent assignment of adverse events, treatment-emergent adverse events (TEAEs) were assigned to the treatment a patient was receiving at the time of each event listed in Table 2. The denominator for each treatment group includes the total of all patients who ever received a particular treatment at any time during the study. The most frequent TEAE was elevations in liver-associated enzymes, including ALP, GGT, and serum aminotransferases, all of which occurred while patients were receiving placebo (14 [40%]). This was believed to be due to withdrawal of treatment, either UDCA itself during period 1 or HTD1801 in period 3. Diarrhea was the most frequent TEAE noted with HTD1801 treatment and occurred in 7 patients (23%) receiving HTD1801 1,000 mg BID.

Table 2.:
Treatment-emergent adverse events occurring in 2 or more patients (safety population)

TEAEs leading to discontinuation of the study drug occurred in 8 patients (15%) as presented in Table 3. The most frequent adverse event leading to discontinuation of the study drug was elevations in levels of liver-associated enzymes, with 9 events occurring in 4 patients all while receiving placebo. The study blind was broken for 2 of these patients while experiencing elevations in liver-associated enzymes during the third period of this study. These 2 patients were on placebo at the time and the liver enzyme elevations were attributed to withdrawal of HTD1801, which they had been on in the immediately previous period. There were no clear trends of TEAEs leading to discontinuation in patients receiving HTD1801.

Table 3.:
Treatment-emergent adverse events leading to study drug discontinuation (safety population)

Four serious adverse events occurred during the course of this study, affecting 4 patients. All 4 events were recorded while patients were receiving HTD1801 (3 on 1,000 mg BID and 1 on 500 mg BID). These SAEs included bleeding from an existing colostomy, partial small bowel obstruction, Clostridioides difficile colitis, and physical injury associated with a motor vehicle accident. None of these were attributed to the study drug by the local investigator.

Regarding patients with underlying inflammatory bowel disease, there were no significant changes in serum levels of C-reactive protein or fecal calprotectin during therapy. Similarly, there was no overall change in the partial Mayo Score. A summary of concomitant medications that patients received with disease-modifying properties for treatment of IBD is presented in Supplemental Table 1, Supplementary Digital Content 1, https://links.lww.com/AJG/C650. Each of these medications was allowed per protocol and was unlikely to cause any drug-drug interactions based on the reported metabolism of the individual compounds. Mesalazine was used by 15 patients with IBD, and budesonide and vedolizumab were used by 3 patients each. All other concomitant medications for IBD were used by 2 or fewer patients.


Change in serum ALP has long been used as an initial end point in clinical trials of novel therapeutic agents for PSC because reductions in ALP have been associated with better clinical outcomes (15–19). This phase 2 proof-of-concept study has shown that HTD1801 is active and effective in PSC by causing substantial reductions in serum ALP. In addition to improvements in ALP, decreases were also observed in other liver-associated enzymes including ALT, AST, and GGT further supporting the beneficial activity of HTD1801.

One of the significant challenges in studying HTD1801 in PSC is that it delivers UDCA. Other compounds in the development of PSC may not necessarily require discontinuation of UDCA or may be intended as an “add-on” therapy. Discontinuing UDCA in patients with PSC is known to result in elevations in ALP with a flare of the disease, which can sometimes be severe (20). This study had to deal with prior UDCA usage in several ways. Some patients stopped their UDCA at least 1 month before entering the study, but most of those who had been on UDCA discontinued only 1 day before starting the study drug. Most of those who had been on UDCA and who were randomized to placebo in the initial treatment period went through the previously well-described flare of disease activity, with elevations in all liver-associated enzymes, including ALP and GGT. At the same time, patients receiving HTD1801 had a reduction in liver-associated enzymes, so to be able to accurately and appropriately compare active treatment with placebo, a planned subset analysis was conducted, separating those who had been on UDCA (regardless of time of UDCA discontinuation) from those with no prior UDCA usage. As shown in Figure 4, for those who had not previously been on UDCA, there was a clear and obvious beneficial effect of HTD1801 because ALP levels decreased on active treatment and were unchanged on placebo. By contrast, for those who had been on UDCA, ALP levels on average remained suppressed when receiving HTD1801 but became elevated when on placebo (which functioned as a form of UDCA withdrawal). As a result, it is likely that some of the difference observed between treatment and placebo overall for changes in ALP is caused by the withdrawal of UDCA. Nonetheless, mean values of ALP declined from baseline overall.

HTD1801 is a novel small molecule, an ionic salt formed between a base (BBR) and a weak acid (UDCA) with a stoichiometry of 1:1. The novel crystalline structure of HTD1801 enhances the solubility of BBR, thus increasing its bioavailability. When HTD1801 is ingested, it dissociates into BBR and UDCA, both of which are then differentially absorbed (8). As noted previously, the actions of UDCA in PSC are well known and may be enhanced when it is delivered in the form of HTD1801 (5–7). BBR has both antimicrobial and anti-inflammatory properties and affects the gut-liver axis by modulating the gut microbiome (21–26). Thus, some of its actions are believed to be systemic while others are believed to be local within the gastrointestinal tract. For example, BBR is widely used in some countries for gastrointestinal problems such as traveler's diarrhea, where its benefits are likely because of the antimicrobial effect of BBR (10). Furthermore, BBR has also been effective against Klebsiella pneumoniae, a commensal organism in the gastrointestinal tract that is believed to play a pivotal role in the pathogenesis of PSC (4). In sum, these properties may play a role in improving the prognosis of PSC.

UDCA has a long history of use in PSC (27–29). It is widely used and effective in primary biliary cholangitis, but in studies of PSC, UDCA has not been of clinical benefit, although significant reductions in ALP are commonly seen (6). This has led to the widespread clinical use of UDCA in patients with PSC, although it is not FDA-approved for the treatment of PSC and at least 1 study has shown that high doses of UDCA may be harmful. A study by Lindor and colleagues of high dose UDCA evaluated 28–30 mg/kg and found that this dose was associated with an unexpectedly higher rate of liver-related death (7). HTD1801 is not administered on a weight-based dose, but as a standard, single dose. The average body weight of patients in this study was about 80 kg. A 2000-mg per day dose (1,000 mg BID) of HTD1801 delivers approximately 1,075 mg of UDCA, so for the average patient in this study, their dose of UDCA would have been approximately 12.5 mg/kg, far less than the amounts delivered when dosing UDCA at 28–30 mg/kg.

This study had several limitations that may influence the interpretation of the results. This was a placebo-controlled study with no active comparison with an effective treatment nor with the constituents of HTD1801, UDCA, and BBR. Therefore, based on the current study results, it is not possible to discern whether the biochemical effects (reduction in ALP and other liver enzymes) observed in this study can be solely attributed to the unique properties of HTD1801 or because of the properties of UDCA or BBR individually. In addition, the protocol amendment allowing for the discontinuation of UDCA immediately before receiving the study drug may have influenced the interpretation of the primary end point as previously described. Future studies may consider a full washout of UDCA with stabilized liver biochemistry before randomization to reduce the confounding effects related to a postbaseline UDCA withdrawal flare as was done in the Phase 2 study of norUDCA (30). This study had a relatively small sample size (N = 55) with a short duration of treatment, the longest period on active treatment being 18 weeks and the primary end point being measured at 6 weeks. Nonetheless, a prompt reduction in ALP was seen as early as 2 weeks. By contrast, in a phase 2 study of obeticholic acid in PSC, serum ALP reached its nadir some time between week 6 and week 12 (31). In the phase 2 study of cilofexor in PSC, serum ALP reached its nadir at week 4 with the 100-mg dose and at week 8 with the 30-mg dose (32). While ALP data seem to have been collected at week 2 in the study of norUDCA, these data are not shown (30).

This study of HTD1801 has demonstrated the proof of principle that it has activity in PSC by showing that its use decreases serum ALP levels. To confirm that HTD1801 is effective in the treatment of PSC, HTD1801 will need to be administered for a longer duration and studies will either need to demonstrate clinical benefit and that it halts or reverses the progression of liver disease or show an effect on other clinically important biomarkers, such as liver stiffness and serum markers of fibrosis. Future studies will also need to further examine the mechanisms of action of HTD1801, including detailed study of changes in the gut microbiome associated with its use, which were not evaluated in this study.

HTD1801 seems to be safe and generally well tolerated at the doses studied. The most common TEAE with HTD1801 treatment was diarrhea, which was mild to moderate in severity. There was no apparent change in status of underlying IBD in patients in this study. In addition, TEAEs of elevations in liver enzymes were observed throughout this study but only occurred in patients in the placebo group and were attributed to withdrawal of active therapy (either UDCA at study initiation or HTD1801 in various study periods).


In this randomized placebo-controlled study of HTD1801, a significant beneficial effect was noted in serum levels of liver-associated enzymes, including ALP. Withdrawal of HTD1801 was associated with marked exacerbations of ALP levels, also confirming the beneficial effect of HTD1801. Finally, HTD1801 was generally well tolerated in this population of patients with numerous comorbidities. Given the known benefits of UDCA in PSC and the broad-spectrum antimicrobial and anti-inflammatory activity of BBR, HTD1801 can be expected to both mitigate the hepatocellular injury associated with PSC and suppress the bacterially driven process that originates PSC and continues to drive it. However, it may require longer term studies to confirm the benefit of the BBR component of HTD1801. These data support further clinical trials of prolonged therapy with HTD1801 to assess the impact of this new molecular entity in preventing clinical events in patients with PSC.


Guarantor of the article: Adrian M. Di Bisceglie, MD.

Specific author contributions: K.V.K., L.F., G.H., and A.M.D. served as a publications committee. B.E., N.G., V.S., C.L., M.P., S.T., A.K., A.L., and C.L. contributed to collection of the data and critical review of the data and manuscript.

Financial support: This study was funded by HighTide Therapeutics, Rockville, MD.

Potential competing interests: K.V.K.: Grant/research support: Gilead, Intercept, Genfit, Enanta, Allergan (now AbbVie), HighTide, CymaBay, NGM BioPharma, Mirum, Madrigal, Pfizer, Viking, GSK, and Terns. Advisory committee or review panel: Gilead, Akero, Intercept, Genfit, Enanta, and NGM BioPharma. Consulting: Calliditas, Intercept, HighTide, Asembly, Protagonist Therapeutics, Mirum, and Madrigal. Speaking and teaching: AbbVie, Gilead, and Intercept. Stock shareholder: Inipharm and Assembly. L.F.: Nothing to disclose. B.E.: Nothing to disclose. N.G.: Grant/research support: Pfizer, Gilead, Inventiva, Sagimet, 89Bio, Axcella, Helio, and Mirum. Speaking and teaching: Gilead. V.S.: Speaking and teaching: Gilead and AbbVie. C.L.: Nothing to disclose. S.A.H.: Grant research: Akero Therapeutics, Axcella Health, Cirius Therapeutics, CiVi Biopharma, CymaBay Therapeutics, Enyo Pharma S.A, Galectin Therapeutics, Galmed Research & Dev., Genfit Corp, Gilead Sciences, Hepion Pharmaceuticals, HighTide Therapeutics, Intercept Pharmaceuticals, Madrigal Pharmaceuticals, Metacrine, NGM Biopharmaceuticals, Northsea Therapeutics, Novartis Pharmaceuticals Corp, Novo Nordisk, Poxel, Sagimet Biosciences, and Viking Therapeutics. Consultant: AgomAB, Akero Therapeutics, Alentis Therapeutics AG, Alimentiv, Altimmune, Axcella Health, Boston Pharmaceuticals, B Riley FBR, BVF Partners LP, Cohbar, Canfite, Corcept Therapeutics, CymaBay Therapeutics, Echosens North America, Enyo Pharma S.A, Fibronostics, Foresite Labs, LLC, Fortress Biotech, Galectin Therapeutics, Genfit Corp, GNS, Hepion Pharmaceuticals, HighTide Therapeutics, HistoIndex PTE, Inipharm, Intercept Pharmaceuticals, Ionis, Kowa Research Institute, Madrigal Pharmaceuticals, Medpace, Metacrine, Microba, NGM Biopharmaceuticals, Northsea Therapeutics B.V, Novo Nordisk, Nutrasource, Perspectum Diagnostics, Piper Sandler, Poxel, Prometic, Pharma SMT, Ridgeline, Sagimet Biosciences, Sonic Incytes Medical Corp, Terns, and Viking Therapeutics. Advisory board/panel: 89 Bio, Akero Therapeutics, Altimmune, Arrowhead, Axcella Health, Chronwell, CiVi, CymaBay Therapeutics, Echosens North America, Foresite Labs, LLC, Galectin Therapeutics, Galmed Research & Dev., Genfit Corp, Gilead Sciences, Hepion Pharmaceuticals, HighTide Therapeutics, HistoIndex PTE, Indalo, Intercept Pharmaceuticals, Madrigal Pharmaceuticals, Medpace, Metacrine, NGM Biopharmaceuticals, Northsea Therapeutics B.V, Novartis Pharmaceuticals, Novo Nordisk, PathAI, Poxel, Prometic Pharma SMT, Ridgeline, Sagimet Biosciences, Sonic Incytes Medical Corp, Terns, and Theratechnologies. Stock/shares (self-managed): Akero Therapeutics, Chronwell, Cirius Therapeutics, Galectin Therapeutics, Genfit Corp, Hepion Pharmaceuticals, HistoIndex PTE, Metacrine, NGM Biopharmaceuticals., Northsea Therapeutics B.V, and Sonic Incytes Medical Corp. C.L.: Grant/research support: Mirum, GSK, Genfit, Pliant, Gilead, Cara Therapeutics, TARGET PharmaSolutions, Shire, NGM, Intercept, Mitsubishi, Alnylam, Novartis, CymaBay, Durect, HighTide, Enanta, Zydus, and Genkyotex. Consulting: Escient, CymaBay, Genfit, GSK, Mirum, Cara Therapeutics, TARGET PharmaSolutions, Shire, Pliant, Intercept, Calliditas, and Gilead. Board membership: ABIM committee member, Associate Editor for Hepatology. A.L.: Employment: HighTide. Stock: Intercept. A.M.D.: Consulting: HighTide. G.M.H.: Consulting: Genfit, Roche, Morphic, Falk, Mirum, Pliant, and CymaBay. Advisory committee or review panel: Intercept and GSK.

ClinicalTrials.gov Identifier: NCT03333928.

Study Highlights


  • ✓ There are no approved pharmaceutical therapies for primary sclerosing cholangitis (PSC).
  • ✓ Berberine ursodeoxycholate (HTD1801) is currently being investigated for the treatment of PSC.


  • ✓ Patients with PSC treated with HTD1801 experienced significant improvement in alkaline phosphatase, a marker of PSC disease severity.
  • ✓ HTD1801 was generally safe and well tolerated.


We acknowledge the contributions of all the patients participating in this study, along with each of the study investigators and site staff. We thank Joe Hirman, PhD, and Abigail Flyer of Pacific Northwest Statistical Consulting for providing statistical and programming support. This study was sponsored by HighTide Therapeutics.


1. Chapman R, Fevery J, Kalloo A, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology 2010;51(2):660–78.
2. Bambha K, Kim WR, Talwalkar J, et al. Incidence, clinical spectrum, and outcomes of primary sclerosing cholangitis in a United States community. Gastroenterology 2003;125(5):1364–9.
3. Karlsen TH. Primary sclerosing cholangitis: 50 years of a gut–liver relationship and still no love? Gut 2016;65(10):1579–81.
4. Nakamoto N, Sasaki N, Aoki R, et al. Gut pathobionts underlie intestinal barrier dysfunction and liver T helper 17 cell immune response in primary sclerosing cholangitis. Nat Microbiol 2019;4(3):492–503.
5. Olsson R, Boberg KM, de Muckadell OS, et al. High-dose ursodeoxycholic acid in primary sclerosing cholangitis: A 5-year multicenter, randomized, controlled study. Gastroenterol 2005;129(5):1464–72.
6. Lindor KD. Ursodiol for primary sclerosing cholangitis. N Engl J Med 1997;336(10):691–5.
7. Lindor KD, Kowdley KV, Luketic VA, et al. High-dose ursodeoxycholic acid for the treatment of primary sclerosing cholangitis. Hepatology 2009;50(3):808–14.
8. Di Bisceglie AM, Watts GF, Lavin P, et al. Pharmacokinetics and pharmacodynamics of HTD1801 (berberine ursodeoxycholate, BUDCA) in patients with hyperlipidemia. Lipids Health Dis 2020;19(1):1–10.
9. Hofmann A. Pharmacology of ursodeoxycholic acid, an enterohepatic drug. Scand J Gastroenterol 1994;29(suppl 204):1–15.
10. Kumar A, Ekavali, Chopra K, et al. Current knowledge and pharmacological profile of berberine: An update. Eur J Pharmacol 2015;761:288–97.
11. Voigtländer T, Leuchs E, Vonberg RP, et al. Microbiological analysis of bile and its impact in critically ill patients with secondary sclerosing cholangitis. J Infect 2015;70(5):483–90.
12. Wang Y, Xiang D, Chen W, et al. Berberine attenuates cholestatic liver and bile duct injury in Mdr2−/− mice by maintaining bile acid homeostasis. FASEB J 2020;34(S1):1.
13. Naghibi Z, Rakhshandeh H, Jarahi L, et al. Evaluation of the effects of additional therapy with berberis vulgaris oxymel in patients with refractory primary sclerosing cholangitis and primary biliary cholangitis: A quasi-experimental study. Avicenna J Phytomed 2021;11(2):154–67.
14. Lewis JD, Chuai S, Nessel L, et al. Use of the noninvasive components of the Mayo score to assess clinical response in ulcerative colitis. Inflamm Bowel Dis 2008;14(12):1660–6.
15. Stanich PP, Björnsson E, Gossard AA, et al. Alkaline phosphatase normalization is associated with better prognosis in primary sclerosing cholangitis. Dig Liv Dis 2011;43(4):309–13.
16. Al Mamari S, Djordjevic J, Halliday JS, et al. Improvement of serum alkaline phosphatase to <1.5 upper limit of normal predicts better outcome and reduced risk of cholangiocarcinoma in primary sclerosing cholangitis. J Hepatol 2013;58(2):329–34.
17. Lindström L, Hultcrantz R, Boberg KM, et al. Association between reduced levels of alkaline phosphatase and survival times of patients with primary sclerosing cholangitis. Clin Gastroenterol Hepatol 2013;11(7):841–6.
18. Rupp C, Rössler A, Halibasic E, et al. Reduction in alkaline phosphatase is associated with longer survival in primary sclerosing cholangitis, independent of dominant stenosis. Aliment Pharmacol Ther 2014;40(11-12):1292–301.
19. Ponsioen CY, Chapman RW, Chazouillères O, et al. Surrogate endpoints for clinical trials in primary sclerosing cholangitis: Review and results from an international PSC study group consensus process. Hepatology 2016;63(4):1357–67.
20. Wunsch E, Trottier J, Milkiewicz MR, et al. Prospective evaluation of ursodeoxycholic acid withdrawal in patients with primary sclerosing cholangitis. Hepatology 2014;60(3):931–40.
21. Yue M, Tao Y, Fang Y, et al. The gut microbiota modulator berberine ameliorates collagen‐induced arthritis in rats by facilitating the generation of butyrate and adjusting the intestinal hypoxia and nitrate supply. FASEB J 2019;33(11):12311–23.
22. Čerňáková M, Košťálová D. Antimicrobial activity of berberine—a constituent of Mahonia aquifolium. Folia Microbiol 2002;47(4):375–8.
23. Wultańska D, Piotrowski M, Pituch H. The effect of berberine chloride and/or its combination with vancomycin on the growth, biofilm formation, and motility of Clostridioides difficile. Eur J Clin Microbiol Infect Dis 2020;39(7):1391–9.
24. Zhang Y, Gu Y, Ren H, et al. Gut microbiome-related effects of berberine and probiotics on type 2 diabetes (the PREMOTE study). Nat Commun 2020;11(1):1–12.
25. Zhang X, Zhao Y, Zhang M, et al. Structural changes of gut microbiota during berberine-mediated prevention of obesity and insulin resistance in high-fat diet-fed rats. PLoS One 2012(8):e42529.
26. Ehteshamfar SM, Akhbari M, Afshari JT, et al. Anti-inflammatory and immune-modulatory impacts of berberine on activation of autoreactive T cells in autoimmune inflammation. J Cell Mol Med 2020;24(23):13573–88.
27. Stiehl A. Ursodeoxycholic acid therapy in treatment of primary sclerosing cholangitis. Scand J Gastroenterol 1994;29(suppl 204):59–61.
28. Mitchell SA, Bansi DS, Hunt N, et al. A preliminary trial of high-dose ursodeoxycholic acid in primary sclerosing cholangitis. Gastroenterol 2001;121(4):900–7.
29. Beuers U, Spengler U, Kruis W, et al. Ursodeoxycholic acid for treatment of primary sclerosing cholangitis: A placebo-controlled trial. Hepatology 1992;16(3):707–14.
30. Fickert P, Hirschfield GM, Denk G, et al. norUrsodeoxycholic acid improves cholestasis in primary sclerosing cholangitis. J Hepatol 2017;67(3):549–58.
31. Kowdley KV, Vuppalanchi R, Levy C, et al. A randomized, placebo-controlled, phase II study of obeticholic acid for primary sclerosing cholangitis. J Hepatol 2020;73(1):94–101.
32. Trauner M, Gulamhusein A, Hameed B, et al. The nonsteroidal farnesoid X receptor agonist cilofexor (GS-9674) improves markers of cholestasis and liver injury in patients with primary sclerosing cholangitis. Hepatology 2019;70(3):788–801.

Supplemental Digital Content

Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of The American College of Gastroenterology