The history and future of bile acid therapies

Summary

Animal bile has been regarded as a potential remedy for liver diseases for millennia in ancient Eastern cultures. The application of bile acids, as major organic components of bile, for gallstone dissolution and of ursodeoxycholic acid (UDCA) for cholestatic liver diseases began in the second half of the 20^th century. Currently, UDCA is regarded as an effective first-line treatment for primary biliary cholangitis, the most common chronic cholestatic liver disease, and is also applied in other cholestatic disorders, including primary sclerosing cholangitis and intrahepatic cholestasis of pregnancy. For the past two decades, nuclear receptor agonists – such as the farnesoid X receptor agonist and the chenodeoxycholic acid derivative obeticholic acid – as well as derivatives of UDCA, such as norucholic acid and various UDCA hybrids, have been under investigation for the treatment of a wide range of benign and malignant, genetic and acquired hepatobiliary disorders. The development of new drugs that interact with the enterohepatic bile acid circulation has attracted attention for rare genetic paediatric disorders. In the present review, we describe the current status and future prospects of bile acid-based therapies for hepatobiliary diseases in adults, children, and during pregnancy, on the basis of historical developments over centuries and millennia.

History of bile acid therapy

About 4,000 years ago, Babylonian and Assyrian priests and diviners in ancient Mesopotamia (modern day Iraq) were the first documented to use liver and gallbladder inspection (hepatoscopy) of sacrificed animals as a tool to foresee future events, such as the health prognoses of prominent individuals or the outcome of decisive battles (Fig. 1).¹ Their habit of hepatoscopy was taken over by Greek, Etruscan and Roman haruspices. While the liver was widely seen as the seat of the soul, the form and circumference of the gallbladder as the reservoir of ‘yellow bile’ was used in the Roman empire to predict the outcome of future events, particularly of upcoming (sea) battles, and were associated with Neptune, the god of the sea. Before the start of the sea battle of Actium (31 BCE.), Octavianus, the adopted son of Julius Caesar, was presented a double gallbladder at hepatoscopy by his haruspices and, as predicted, won the famous battle against Marcus Antonius to become Emperor Augustus.¹ Hepatoscopy was practiced until the end of the Roman empire, although intellectual leaders like the Greek philosopher and scientist Aristotle or the Roman philosopher and politician Cicero seriously questioned this habit and the intellectual capacity of those who believed in it.¹

Fig. 1. Babylonian Terracotta model of a sheep liver with a prominent gallbladder and cystic duct as reservoir of ‘yellow bile’ and covered with cuneiform characters in Akkadian language saying: ‘May your liver be smooth’.

~1600-2000 BCE. British Museum, London.

In the ancient Egyptian Ebers papyrus (~1550 BCE), a medical summary of about 700 diseases on 110 papyrus pages, bile was first mentioned as a potentially useful remedy and purge.¹ An ancient medical anthology from a tomb in Southern China dated back to about 300 BCE., recommended the use of animal bile for jaundiced patients based on the therapeutic principle Yi Du Gong Du, that is, combatting poison with poison. ² Strikingly, bear bile was already considered the “king” of animal biles in that anthology. In the Chinese Tang Materia Medica, the worldwide first state pharmacopoeia collected during the classical period of Chinese art and literature of the T’ang dynasty (618-907 CE.) it was stated that “bear bile which tastes bitter can be used to treat jaundice, … intractable diarrhea in summer and cardialgia”.³ In Eastern Asian societies, the belief in a therapeutic effect of bile constituents has persisted for millennia and dried powder of bear bile collected on bear farms and available in drug stores is still used today by some practitioners of traditional Chinese medicine as a remedy for liver diseases.³

The Western world’s view changed more than two millennia ago with bile components ignored as potential remedies for most of the intervening period. The Corpus Hippocraticum, a collection of Ancient Greek medical texts, research notes and philosophical reflections, named after Hippocrates of Kos (460-377 BCE.), suggested that a dysbalance (dyscrasia) of the – in their view – four cardinal humours of the human body, phlegm (brain), yellow bile (liver), blood (heart) and black bile (spleen), may be a frequent cause of internal diseases.¹ The Greek philosopher and scientist Aristotle who compared the anatomy of the gallbladder and bile ducts in various animal species regarded bile and faeces as useless excreta.^1,4 The physician scientist Erasistratus of Alexandria, the centre of medical and biologic research of the Hellenistic period, who offered the first detailed hypothesis on bile formation and excretion, shared the view of bile as a useless and potentially harmful fluid. Finally, Galen of Pergamon (~130-201 CE.), the most prolific physician and researcher of the ancient world, evidenced by the size of his body of work, concluded that yellow bile, the light, foamy and bitter component of chylus, was sucked into the gallbladder and poured out into the intestine to stimulate intestinal contraction and defecation, clean the intestinal interior and dye the feces.¹ He did not consider bile components as potential remedies. His medical writing became dogma for 1500 years as did the early Christian literature in the Western world when a dogmatic Christian Church stifled all creative scientific thinking during the Middle ages.^4,5

The re-establishment of scientific thinking by Leonardo da Vinci, Andreas Vesalius and others in the late 15^th and 16^th century included extensive studies on hepatobiliary anatomy and physiology, sparking growing interest in bile and its components.^4,5 An enterohepatic circulation of bile components was first proposed by Reverhorst (1690) and Borelli (1732).^4,5 The ‘Elementa Physiologica Corporis Humani’ of von Haller (1764), the first synthesis of anatomy and physiology for 1600 years,⁵ stated that ‘bile has the properties of soap when it is mixed with water and it solves fat in water’.⁴ First chemical bile analyses by various investigators identified bile constituents like bile salt crystals (Ramsay, 1757), cholesterol crystals (Conradi, 1775), bile salts and bilirubin (Fourcroy, 1801; Thenard, 1806, Berzelius, 1808) in an alkaline soap-like watery fluid.⁵ The chemists Gmelin, Demarcay and von Liebig first purified different bile acids in the 1820s and 1830s; von Liebig introduced the term ‘bile acid’ (BA) in 1843, as well as developing artificial nitrate-based manure/fertilizer and the modern mirror – three items we are still familiar with in the 21^st century.^4,5

The development of analytical colorimetric methods enabled isolation and elucidation of the approximate structure of cholesterol and several primary (glycine and taurine conjugates of cholic acid (CA) and chenodeoxycholic acid (CDCA)) and secondary (deoxycholic acid (DCA) and lithocholic acid (LCA)) human BAs 80 years later by Windaus and Wieland who were awarded the Nobel Prize in 1928 and 1929 for their pioneering work. The Japanese chemists Shoda (1927) and Iwasaki (1936) were the first to isolate and structurally characterise the human tertiary BA ursodeoxycholic acid (UDCA) from bear bile.⁴ Subsequent studies demonstrated that in black bears, UDCA primarily exists in its taurine-conjugated form, tauroursodeoxycholic acid (TUDCA).⁶ TUDCA was found to comprise up to 62% of the total BAs in American black bear bile (range: 0–62%, mean: 34%, n = 75) and 59% in a sample of dried Chinese black bear bile (n = 1).^3,6

UDCA: from bear bile to post-translational secretagogue

UDCA, 3α,7β-dihydroxy-5β-cholanoic acid, is regarded as a tertiary human BA, amounting for 1–3% of human BAs. As mentioned, UDCA has been administered as a remedy for liver diseases in the form of dried black bear’s bile in traditional Chinese medicine for more than two millenia.² Enrichment of UDCA in black bear’s bile is similar to that of individuals with cholestatic liver diseases under treatment with UDCA at recommended therapeutic doses of about 15 mg/kg/day.⁷ In Japan, UDCA was launched in the late 1950s by Tokyo Tanabe, making its subsequent evaluation as a gallstone-dissolving (see below) and later anticholestatic drug possible.

Primary biliary cholangitis

In the Western literature, the beneficial effects of UDCA on serum liver tests in individuals with primary biliary cholangitis (PBC) were first reported in peer-reviewed form in the late 1980s by Poupon’s and Leuschner’s groups from Paris and Frankfurt.^8-10 Large randomised, placebo-controlled trials confirmed these preliminary observations and suggested that UDCA not only improves biochemical markers of cholestasis and liver injury in blood, but also reduces inflammatory activity in the liver, halts liver fibrosis and appears to improve liver transplant (LT)-free survival in PBC, leading to normal life expectancy under long-term treatment with adequate doses of 13-15 mg/kg/day in two thirds of treated individuals.^11-15 Today, UDCA (13-15 mg/kg/day) is the standard first-line treatment for PBC worldwide on the basis of prospective and retrospective data obtained over 30 years by global networks (Fig. 2; Table 1).^15-18

Fig. 2. UDCA improves transplant-free survival in complete and ‘incomplete’ responders with primary biliary cholangitis (GLOBE score).¹⁵.

HR, hazard ratio; LT, liver transplant; UDCA, ursodeoxycholic acid.

Table 1.

BA and BA-related therapies, targeted diseases, recommended doses, evidence (Oxford Centre for Evidence-based Medicine) and recent guidelines.

BA	Disease	Recommended dose	Guideline recommendation	Evidence	References (Guidelines)
UDCA	PBC	13-15 mg/kg/day	EASL recommends oral UDCA at 13–15 mg/kg/day as the first-line pharmacotherapy for all patients with PBC. UDCA is usually continued for life.	LoE 1 (strong recommendation, strong consensus)	16,18,239
UDCA	PBC post-transplant	13-15 mg/kg/day	EASL suggests that in patients with proven or likely recurrent PBC post LT, the use of UDCA is safe and can improve liver biochemistry.	LoE 2 (see text for data supporting use of UDCA post LT in all PBC)	16,25
UDCA	PSC	15(−20) mg/kg/day	UDCA at doses of 15-20 mg/kg/day can be given since it may improve serum liver tests and surrogate markers of prognosis.	LoE 2 (weak recommendation, 76% consensus)	26,35
UDCA	CCA in PSC	15 mg/kg/day	—	Very limited	26,35
UDCA	Colon cancer in PSC-IBD	≥15 mg/kg/day	—	Limited	26
UDCA	ICP	10-20 mg/kg/day	UDCA should be considered for treatment of maternal pruritus … In ICP and sBA concentrations ≥40 mmol/L, UDCA should be offered as a treatment to reduce the risk of spontaneous preterm birth and it may also be protective against stillbirth.	LoE 2 (strong recommendation, strong consensus) LoE 2 (strong recommendation, consensus)	205
UDCA	ABCB4/MDR3 deficiency (PFIC3)		Recommended for those with at least one ABCB4 missense variant and a clinical phenotype. Partial response to UDCA has been reported in patients carrying biallelic protein-truncating MDR3/ABCB4 gene variants and UDCA may therefore be offered. Long-term UDCA therapy is recommended for recipients of an MDR3 monoallelic deficient graft.	LoE 3 (strong recommendation, strong consensus) LoE 4, weak recommendation, strong consensus) LoE 4 (strong recommendation, strong consensus)	166
UDCA	Gallstone prevention	≥500 mg/day	In situations that are associated with rapid weight loss (e.g. very-low-calorie diet, bariatric surgery), temporary UDCA may be recommended.	LoE 2 (moderate evidence; weak recommendation)	240
OCA	PBC (2nd line)	Start 5 mg/day, max. 10 mg/day	OCA had been conditionally approved for patients with PBC in combination with UDCA for those with an inadequate response to UDCA, or as monotherapy in those intolerant to UDCA. Full approval was recently denied by American and European authorities.	See text for recent developments	16
NCA	PSC	1,500 mg/day		(Phase III results awaited, promising preliminary data)	84
IBAT/ASBT inhibitors	Alagille syndrome	(Depends on substance)	When available, IBAT inhibitor treatment should be offered to patients with Alagille syndrome and cholestatic pruritus.	LoE 2 (strong recommendation, strong consensus)	166
IBAT/ASBT inhibitors	FIC1/ATP8B1 deficiency (PFIC1)		Medical or surgical interruption of the enterohepatic circulation at transplantation or post-transplant could be used to treat diarrhoea and steatohepatitis.	LoE 5 (weak recommendation, strong consensus)	166
IBAT/ASBT inhibitors	BSEP/ABCB11 deficiency (PFIC2)		Medical interruption of the enterohepatic circulation by IBAT inhibition should be considered to reduce pruritus prior to considering liver transplantation in patients with missense mutations.	LoE 2 (strong recommendation, strong consensus)	166
CA or CDCA	Primary BA synthesis defects		Patients with a primary bile acid synthesis defect (namely HSD3B7 and AKR1D1 deficiencies) should be treated with an oral primary bile acid.	LoE 3 (strong recommendation, strong consensus)	166

AKR1D1, Δ⁴-3-Oxosteroid 5β-reductase; ASBT, apical sodium bile acid transporter; CA, cholic acid; CDCA, chenodeoxycholic acid; HSD3B7, 3β-hydroxy-Δ5-C27-steroid oxidoreductase deficiency; IBAT, ileal bile acid transporter; LoE, level of evidence; NCA, norucholic acid; OCA, obeticholic acid; sBA, serum bile acid; UDCA, ursodeoxycholic acid.

In people with PBC who show an incomplete response to UDCA alone, combination therapies have shown promise in further improving long-term outcomes. These include UDCA combined with the farnesoid X receptor (FXR) agonist obeticholic acid,¹⁹ the pan-peroxisome proliferator-activated receptor (PPAR) agonist bezafibrate,²⁰ the dual PPARα/δ agonist elafibranor,²¹ or the PPARδ agonist seladelpar.²² In the future, a head-to-head comparison of different combination treatments would be desirable.²³ We also need to better understand the role of different PPAR receptors (α, δ (β), γ) on the anticholestatic and anti-inflammatory action of PPAR agonists and to critically re-evaluate the receptor specificity of the drugs currently under study. For the few people living with PBC who show an incomplete response to dual therapy, the option of triple therapy with UDCA + a PPAR agonist + an FXR agonist (and/or ileal bile acid transporter (IBAT) inhibitors?) may become an option in the near future.²⁴

Routine post-transplant use of UDCA is also recommended in PBC.^16-18,25 In the largest post-transplant cohort of people living with PBC studied by the Global PBC study group (n = 780), prophylactic UDCA treatment (10-15 mg/kg/day) was associated with a reduced risk of disease recurrence, graft loss and death.²⁵

Primary sclerosing cholangitis

Routine use of UDCA in primary sclerosing cholangitis (PSC) is more controversial due to a lack of adequate randomised, placebo-controlled, prospective long-term studies at moderate UDCA doses.²⁶ Pilot randomised and observational trials showed clear improvements in biochemical markers and – when evaluated – partial improvements in histological and cholangiographic criteria after 1-2 years of treatment with moderate doses of 13-20 mg/kg/day. However, these studies did not demonstrate a significant effect on clinical outcomes, likely due to their short duration and the limited number of patients included.^27-31 A trial of high-dose UDCA (28-30 mg/kg/day) vs. placebo showed higher rates of complications in the UDCA group, leading to negative recommendations for high-dose UDCA.³² The largest (n = 198 for analysis) and longest (5 years) randomised, placebo-controlled trial in PSC using moderate UDCA doses (17–23 mg/kg/day) showed no significant biochemical response – in contrast to the improvements seen in several earlier short-term trials (1-2 years) with similar^27-31 or higher³² UDCA doses – raising doubts about the compliance of individuals who were only seen once every 6 months in the 5-year-trial.³³ This longer term study only showed a trend towards improved transplant-free survival, but the trial was underpowered with 198 (analysed) of 346 individuals needed (57%) to reach this clinical endpoint, as determined a priori.³³ In a Japanese cohort of 325 individuals with primary sclerosing cholangitis (PSC) followed for a mean of 5.1 years, there were 57 deaths and 24 LTs during the study period. Patients receiving UDCA monotherapy (n = 202) or combination therapy with UDCA and bezafibrate (n = 76) had better LT-free survival compared to those who received no UDCA treatment (n = 45), with an adjusted hazard ratio of 0.47 (95% CI 0.28–0.78, p = 0.003).³⁴ Publication of a retrospective analysis by the International PSC Study Group of a large cohort followed over longer time intervals – like that published by the Global PBC Study Group for PBC (n = 3,902)¹⁵ – is still awaited, but may provide additional insights on the optimal use and dosage of UDCA in PSC. The 2022 EASL Clinical Practice Guidelines on sclerosing cholangitis recommended, based on available evidence, that “UDCA at doses of 15-20 mg/kg/d can be given since it may improve serum liver tests and surrogate markers of prognosis. Available data does not allow for a firmer recommendation.”²⁶ The 2023 AASLD Practice Guidance recommended that “all patients with PSC should be considered for participation in clinical trials; however, ursodeoxycholic acid (13–23 mg/kg/day) can be considered and continued if well tolerated with a meaningful improvement in alkaline phosphatase (γ-glutamyl transferase in children) and/or symptoms with 12 months of treatment.”³⁵ Finally, it has been shown that stopping UDCA treatment in adult individuals with PSC can lead to worsening of clinical symptoms and biochemical tests, which improve again after reinstitution of UDCA therapy^26,36 (Table 1).

UDCA and malignancies of the hepatobiliary tract in PSC

Indirect evidence had suggested in some previous analyses that UDCA at low to moderate doses might decrease the risk of biliary malignancies in PSC, as the cholangiocarcinoma (CCA) rates were lower than those reported in untreated individuals with PSC.^37,38 Analysis of the large International PSC Study Group cohort of treated and untreated individuals with PSC will provide more valid information on the use of UDCA at different doses and the risk of CCA in PSC during long-term follow-up.

UDCA and colon carcinoma

Initial reports describing the potential benefits of moderate doses of UDCA on the development of dysplasia or cancer in the colon in people with PSC and accompanying inflammatory bowel disease (IBD)^39,40 were only partially supported by meta-analyses of UDCA doses of 8-15 mg/kg/day⁴¹ or <25 mg/kg/day.^26,42 One report with a sensational title and methodological shortcomings described an increased incidence of low grade dysplasia, but no difference in high grade dysplasia or carcinoma, in the colon after treatment with a very high dose of UDCA (28-30 mg/kg/day) compared to placebo.⁴³

Taken together, the long-term effects of UDCA in PSC and PSC-IBD appear closely related to the dose used, and a daily dose of 15 mg/kg/day appears safe and potentially useful for the long-term treatment of hepatobiliary and intestinal disease in PSC.

The use of UDCA in genetic cholestatic liver diseases, pregnancy-related liver diseases such as intrahepatic cholestasis of pregnancy (ICP) and for the prevention of cholelithiasis during rapid weight loss is discussed below (see also Table 1).

Potential mechanisms of action of UDCA

Anticholestatic effects

Potential mechanisms and sites of action of UDCA in cholestatic liver diseases have been intensely studied since the late 1980s. The initial ‘displacement hypothesis’^8,11 suggesting that the anticholestatic effect of UDCA is caused by displacement of hydrophobic BAs from the human BA pool at therapeutic UDCA doses was not supported by subsequent findings. Indeed, human pool size analyses using stable isotope dilution techniques showed that, despite improving serum liver tests in PBC and PSC, short-term UDCA treatment did not lead to an early decrease in the pool sizes of the major hydrophobic BAs CDCA and DCA,⁴⁴ supporting the view of Kitani^45,46 and others that direct effects of UDCA on hepatocytes and cholangiocytes were more relevant. The first hints that UDCA may act as a potent intracellular signalling molecule in hepatocytes (and cholangiocytes) came from the early 1990s with the demonstration that UDCA is a potent Ca⁺⁺ agonist^47-50 and activator of protein kinase Cα (cPKCα)^51,52(-later confirmed in cholangiocytes⁵³), MAPK⁵⁴ and α5β1 integrins⁵⁵ in hepatocytes. The concept that glycine and taurine conjugates of UDCA might enhance the secretory capacity of hepatocytes (and cholangiocytes) by activating intracellular signalling cascades, promoting vesicular trafficking and fusion with the apical membrane, and thereby facilitating the insertion of key transport proteins – such as the bile salt export pump (BSEP/ABCB11) and the conjugate export pump (MRP2/ABCC2)^56-58 – was independently developed by Häussinger et al.⁵⁹ (in the context of bile acid-induced hepatocyte swelling) and Beuers et al.⁵⁰ (as a Ca⁺⁺-/cPKCα-dependent vesicle fusion process, particularly under cholestatic conditions). These groups and others subsequently demonstrated in experimental models that UDCA conjugates exert choleretic and anticholestatic effects by promoting apical carrier insertion via a dual MAPK- and integrin-dependent mechanism in healthy livers⁶⁰ and via Ca⁺⁺/type II inositol-1,4,5-triphosphate receptor/cPKCα/PKA-dependent mechanisms in cholestatic livers.^57,58,61 The concept of a ‘biliary bicarbonate umbrella’ stabilised by UDCA conjugates – as a crucial defence mechanism of human cholangiocytes (and hepatocytes) against the toxic effects of millimolar biliary hydrophobic BAs, particularly GCDCA (pKa ~4) in humans^50,62-64 – further supports the view that the anticholestatic effect of UDCA conjugates in diseases such as PBC, PSC and ICP is mainly explained by their action as post-translational secretagogues in hepatocytes and cholangiocytes (Fig. 3).⁶⁵

Fig. 3. Major putative mechanisms and sites of action of UDCA in cholestatic diseases.⁶⁵.

(A) UDCA conjugates exert anticholestatic effects and stimulate BA and organic anion secretion mainly via post-translational molecular mechanisms (at the level of the hepatocytes) and biliary bicarbonate secretion (at the level of hepatocytes and cholangiocytes), leading to stabilisation of a ‘biliary bicarbonate umbrella’.^62-65 Anti-apoptotic and anti-inflammatory effects are also described.^66-68 (B) The ‘biliary bicarbonate umbrella’ model on the apical membrane of a cholangiocyte: The glycocalyx on the apical membrane may trap bicarbonate molecules in order to create an alkaline milieu which keeps BAs in the negatively loaded state of membrane-impermeable bile salts and, thereby prevents carrier-independent invasion of apolar hydrophobic BAs into cholangiocytes (and hepatocytes). BA, bile acid; UDCA, ursodeoxycholic acid.

Anti-apoptotic effects

The anti-apoptotic effects of UDCA at the level of human and murine hepatocytes were identified and characterised early on.^66-68 However, the impact of the anti-apoptotic effects of UDCA in the clinical setting of cholestatic diseases like PBC and PSC remains unclear at present.³

Immunomodulatory effects

BAs have recently been shown to affect T-cell homeostasis in mice, in particular in liver tumour models. The primary BA CDCA stimulated natural killer T-cell activity and immunosurveillance, whereas the secondary BA LCA and its metabolites contributed to T-cell exhaustion.⁶⁹ Most recently, an immunomodulatory effect of UDCA was tested in experimental mouse models of liver cancer.⁷⁰ Tumour-infiltrating T cells showed BA accumulation, endoplasmic reticulum stress and impaired T-cell function possibly due to impaired secretion of accumulating LCA metabolites.⁷⁰ Improvement of BA and organic anion exporter expression and BA and organic anion secretion under UDCA, a well-known phenomenon in hepatocytes exposed to LCA and UDCA conjugates,^57,58,71 was accompanied by improved tumour-specific T-cell function, enhanced cytokine release and better responsiveness to immunotherapy leading to reduced tumour growth.⁷⁰ Thus, these experimental data support clinical evaluation of UDCA in hepatocellular carcinoma, particularly as an approach to enhance responsiveness to immunotherapy.

NorUDCA (norucholic acid): development and clinical prospects

NorUDCA, recently renamed as norucholic acid (NCA; WHO Drug information 2020;34(3)) is a side chain-shortened derivative of UDCA where the absence of one methyl group confers relative resistance to conjugation with taurine or glycine.^72,73 Originally developed by Alan Hofmann and co-workers to demonstrate the concept of cholehepatic shunting by unconjugated (i.e. conjugation-resistant) BAs resulting in a bicarbonate-rich ‘hypercholeresis‘ (i.e. the amount of bile flow exceeds the amount of administered bile acids),^72,74 NCA was subsequently developed as a therapeutic approach to target diseased bile ducts.⁷⁵ NCA showed profound beneficial effects in several mouse models of cholestatic^75-78 and non-cholestatic^79-81 liver injury, supporting its clinical development. Initial phase II studies have yielded positive results^82,83 and further clinical trials are ongoing in PSC (NCT03872921; phase III; n = 301), with promising preliminary results after 2 years in the presence (78.7%) and absence (21.3%) of concomitant UDCA.⁸⁴ PBC (EudraCT2021-001431-56) and MASLD/MASH (NCT05083390) are among a growing list of other potential therapeutic indications.⁸⁵

According to the cholehepatic shunting hypothesis put forward by A. Hofmann, unconjugated BAs such as NCA are secreted in their anionic form into bile where they are converted into a protonated form, enabling passive absorption by cholangiocytes and generating a bicarbonate anion.^86,87 Notably, induction of bicarbonate secretion – predicted to reinforce the biliary bicarbonate umbrella, protecting cholangiocytes against BA toxicity^62-64 – has also been confirmed in human volunteers.⁷³ Stimulation of a bicarbonate-rich choleresis as a key mechanism to protect cholangiocytes against BA toxicity may become particularly evident in the Mdr2 knockout (KO) mouse model, which lacks biliary phospholipids – another main intrinsic defence system against BA toxicity.⁸⁸ This may at least partly explain the profound therapeutic effects of NCA on sclerosing cholangitis in this model.⁷⁵ Interestingly, further side chain shortening, resulting in bisnorUDCA, aggravated liver injury when administered to Mdr2 KO mice, while generation of taurine-norUDCA resulted in loss of therapeutic properties, emphasising the importance of resistance to side chain conjugation for cholehepatic shunting.⁷⁶ Recent findings support that cholehepatic cycling may not require active transport systems such as the apical sodium-dependent bile acid transporter (ASBT), organic solute transporter-α (OSTα), and organic anion transporting polypeptide 1a/1b (OATP1a/1b)⁸⁹ while the role of the bile salt export pump (BSEP/ABCB11) still remains to be determined. Notably, stimulation of biliary bicarbonate by NCA is independent of CFTR⁷⁶ opening interesting perspectives for the treatment of cystic fibrosis-related liver disease, although initial experimental findings in Cftr KO mice treated with dextran sodium sulfate did not show beneficial effects.⁹⁰ In addition to passive cholehepatic shunting, NCA can also stimulate bicarbonate secretion via activation of calcium-activated chloride channel transmembrane member 16A (also known as anoctamin-1).⁸⁹ Importantly, the therapeutic mechanisms of NCA may also include more direct anti-inflammatory, immunomodulatory and antifibrotic effects in addition to its anticholestatic and choleretic actions.^79,80,91 As such, unconjugated NCA may access immune cells lacking specific BA uptake systems (required for conjugated BAs to pass cell membranes) and alter immunometabolism (e.g. via mTOR inhibition) of CD4⁺ and CD8⁺ T cells, thus attenuating their proliferation and T cell-driven immunopathology^80,91 Notably, such immunomodulatory mechanisms have also been observed in mouse models of intestinal inflammation, where NCA has been shown to suppress Th17 effector function and increase regulatory T-cell abundance by attenuating a glutamine-mTORC1-glycolysis signalling axis.⁹¹ These findings could be clinically relevant for targeting the gut-liver axis in PSC. Moreover, NCA also restores defective TGR5 expression in cholangiocytes, which may contribute to its anti-inflammatory/cholangioprotective effects and to the stimulation of bicarbonate secretion.⁹² Collectively, multiple pleiotropic effects appear to contribute to the therapeutic mechanisms of NCA, which are independent from FXR (as one of the current mainstream approaches in the treatment of cholestasis), while stimulation of PPARγ may be involved.^81,93

The profound therapeutic effects of NCA on sclerosing cholangitis in the Mdr2 KO mouse model⁷⁵ and other non-cholestatic models of liver fibrosis⁷⁹ encouraged clinical studies in PSC, with a phase II study reporting positive effects on biochemical markers of cholestasis and liver injury.⁸² The results of a recently completed phase III study with biochemical, histological and clinical endpoints are eagerly awaited. Apart from ongoing studies in PBC and MASH, further potential clinical indications include ABCB4 deficiency including low phospholipid associated cholelithiasis (LPAC) (mechanistically the human equivalent of the Mdr2/Abcb4 KO mouse model) and the prevention/treatment of non-anastomotic biliary strictures following liver transplantation, where reduced biliary phospholipid excretion and the protective effects of a bicarbonate-rich choleresis may also play an important role.⁸⁵ Further studies will have to explore whether other conjugation-resistant BAs (also shown to undergo cholehepatic shunting and to induce a bicarbonate-rich hypercholeresis)^94,95 share some of the therapeutic properties of NCA.

Novel UDCA derivatives for the treatment of liver diseases

The unique properties of BAs, particularly their specific transport mechanisms and preferential enterohepatic biodistribution, have opened promising avenues for developing novel targeted therapies for liver diseases. Numerous studies have investigated modifications to the chemical structures of natural BAs to create derivatives or hybrid molecules that combine the effects of established drugs with the targeted biodistribution and pharmacokinetic profiles of BAs.^96,97

In this context, UDCA stands out as an ideal scaffold for developing new "Trojan horse" drugs due to its hepatotropism (Fig. 4). Hepatocytes typically conjugate UDCA with taurine or glycine, enabling these residues to serve as sites for attaching bioactive chemical moieties. Additionally, potential deconjugation of these UDCA derivatives by the gut microbiome at the taurine or glycine site during enterohepatic circulation could release more UDCA, providing further benefits due to its hydrophilic, hepatoprotective, and choleretic features.

Fig. 4. General concept of UDCA derivatives and hybrid drugs.

DHA, dihydroartemisinin; HDACi, histone deacetylase inhibitors; TDP1is, tyrosyl-DNA phosphodiesterase 1 inhibitors; UDCA, ursodeoxycholic acid.

UDCA-derived drugs have been explored in gastrointestinal oncology.^98-107 These include hybrids with cisplatin, paclitaxel, nitrogenous bases, histone deacetylase (HDAC) inhibitors, tyrosyl-DNA phosphodiesterase 1 inhibitors, and dihydroartemisinin, among others, all of which have shown promising antitumor effects. These hybrid drugs aim to expand treatment options for hepatobiliary disorders while improving both safety and efficacy. One example is Bamet-UD2,^106,108 which consists of two UDCA molecules linked to cisplatin. This configuration markedly improves hepatobiliary targeting and facilitates the delivery of the chemotherapeutic agent into hepatocellular carcinoma cells via OATP1B3¹⁰⁹ and CCA cells via ASBT.¹¹⁰ This approach enhances antitumor efficacy while reducing toxic side effects compared to cisplatin alone.

A new family of synthetic UDCA derivatives has been developed to selectively inhibit HDAC6 for the treatment of polycystic liver disease (PLD), which currently lacks effective treatments.¹¹¹ HDAC6 overexpression in PLD contributes to ciliary abnormalities and cholangiocyte proliferation.¹¹² However, existing HDAC6 inhibitors lack liver specificity and may cause systemic toxicity due to their aromatic structures.¹¹³ Chronic UDCA treatment has been shown to reduce hepatic cyst growth¹¹⁴ and alleviate symptoms in patients.¹¹⁵ Given its unique structural properties and its ability to conjugate with taurine or glycine, UDCA was explored as a foundation for designing selective HDAC6 inhibitors.¹⁰⁷ Its polycyclic structure allows UDCA to function as a conformationally restricted analogue to macrocyclic systems, enhancing its incorporation into HDAC6 inhibitor designs with improved binding and inhibitory capabilities. By modifying UDCA through the addition of hydrophobic spacers and metal-chelating moieties, four compounds were developed that selectively inhibit HDAC6. Among these, UDCA-HDAC6 inhibitor #1 clearly demonstrated therapeutic efficacy, exhibiting hepatobiliary-targeted pharmacodynamics after oral administration. Importantly, UDCA serves not only as a scaffold but also enhances the pharmacological activity of these inhibitors by improving their binding and selectivity for HDAC6 through increased van der Waals interactions and hydrogen bonding.¹⁰⁷ Preclinical studies indicated that UDCA-HDAC6 inhibitor compounds selectively target the liver, reduce cyst formation, restore ciliary length in cholangiocytes, and suppress cholangiocyte proliferation, all with minimal off-target effects. These HDAC6 inhibitors are effectively transported via BA and organic cation uptake mechanisms, highlighting their potential for treating PLD and other liver disorders. Ongoing research is also exploring selective UDCA-HDAC inhibitors against HDAC1/2 for the treatment of CCA.

BA-derived FXR agonists in hepatobiliary diseases

Given the central role of FXR in maintaining BA homeostasis, as exemplified by Fxr KO mice¹¹⁶ and humans with FXR mutations (PFIC5)¹¹⁷ it appears logical to pursue pharmacological stimulation of FXR as a therapeutic strategy in hepatobiliary disease. Notably, FXR-mediated adaptive changes such as repression of BA uptake systems (NTCP) and BA synthesis (CYP7A1), as well as stimulation of alternative basolateral efflux systems (OSTa/b, MRP4) in response to accumulating BAs, may represent an intrinsic mechanism aimed at restoring BA homeostasis and counteracting BA toxicity in cholestasis.¹¹⁸ However, such adaptive mechanisms are apparently too weak to prevent cholestatic liver injury, perhaps partly due to (post)transcriptional impairment of FXR signalling,¹¹⁹ which may still be amenable to further pharmacological stimulation through highly potent, high affinity FXR agonists. Notably, recent data suggest that BA synthesis may already be maximally repressed in advanced cholestatic liver disease and thus not benefit from further pharmacological suppression via FXR.¹²⁰ However, stimulation of FXR also has anti-inflammatory and antifibrotic effects.¹²¹ Along these lines, recent experimental data obtained in the Mdr2 KO mouse model of sclerosing cholangitis suggest that systemic anti-inflammatory activity may be necessary for full therapeutic effect, which may not be accomplished by targeting BA homeostasis alone.¹²² These findings may explain some of the limitations and negative results seen with gut-restricted FXR agonists in PSC.¹²³

The steroidal FXR agonist OCA is a 6-ethyl derivative of CDCA which showed beneficial effects on prognostically relevant biochemical response criteria (POISE criteria) resulting in conditional approval for the treatment of PBC.¹⁹ Despite initial evidence for long-term efficacy¹²⁴ with improvement of liver histology,¹²⁵ GLOBE and UK-PBC risk scores,¹²⁶ as well as increased transplant-free survival compared to real world external controls,¹²⁷ the confirmatory COBALT trial did not meet its primary composite endpoint, defined as death, LT, model for end-stage liver disease score ≥15, uncontrolled ascites, or hepatic decompensation events (28.6% with OCA and 28.9% with placebo).¹²⁸ In addition to the expected side effect of pruritus, an even higher number of LTs and deaths, as well as a higher risk of drug-induced liver injury (DILI), was observed in the OCA group compared to the placebo group. Although the study was terminated early due to recruitment difficulties (in part associated with post-marketing availability of OCA), comparison with a prespecified external control group showed that OCA treatment was associated with a reduction in death, LT, and hepatic decompensation.¹²⁸ The clinical development of OCA was accompanied by multiple FDA warnings on serious liver injury with incorrect high dosing (September 2017), emphasising the need for correct dosing (February 2018), use restriction in patients with advanced cirrhosis or a history of hepatic decompensation (May 2021) and that serious liver injury could occur in patients without cirrhosis (December 2024). Analyses of all available data by regulatory agencies in both Europe and the United States raised concerns about study methodology, overall safety and efficacy, as well as the risk-benefit ratio, leading to a recommendation for withdrawal from the European market and an FDA advisory committee vote to potentially refrain from full approval for OCA at the end of 2024. Paradoxically, a major threat to efforts to improve treatment in PBC is the regulatory approval model involving conditional marketing authorisation for new drugs based on biochemical markers on the condition that long-term, randomised, placebo-controlled outcome trials are performed to confirm efficacy.¹²⁹ As demonstrated by the COBALT trial, it is difficult to retain patients in the required placebo-controlled PBC outcome trials for up to 5 years when a licensed drug is already available and novel alternative therapies such as PPAR agonists are also available or within reach.¹²⁹

Part of OCA’s toxicity could be related to its steroidal structure and BA properties as a 6-ethyl derivative of CDCA, which is a hydrophobic and toxic BA. Notably, OCA had no beneficial effects in the Mdr2 KO mouse model of sclerosing cholangitis lacking biliary phospholipid excretion, which represents a scenario expected to aggravate any potential toxic BA effects.¹³⁰ In the same study, the FXR/TGR5 dual agonist INT-767 had beneficial therapeutic effects, notably in part due to transcriptional stimulation of the bicarbonate secretory machinery, which were demonstrated to be mediated exclusively by FXR and not TGR5.¹³⁰ Nevertheless, OCA (metabolites) may also activate TGR5,¹³¹ which has been linked to the pathogenesis of pruritus^132,133 and therefore could be implicated in the side effects of OCA. It will be interesting to see whether follow-up compounds such as INT-787 also show beneficial effects in cholestasis.¹³⁴

The development of non-steroidal FXR agonists, several of which show a gut preferential action and shorter hepatic/systemic exposure, initially raised expectations for a better safety profile, with less severe pruritus and dyslipidaemia, than seen with OCA.^135,136 Cilofexor improved cholestatic liver enzymes in individuals with PSC.^137,138 However, a phase III study in non-cirrhotic people with PSC has been discontinued because of the lack of antifibrotic effects,¹³⁹ despite encouraging preclinical effects in the Mdr2 KO mouse model.¹⁴⁰ Tropifexor improved gamma-glutamyltransferase but not alkaline phosphatase levels in people with PBC, which might be attributed to high FXR affinity-related activation of ALP transcription.¹⁴¹ Notably, pruritus and dyslipidaemia were also observed with non-steroidal FXR ligands¹³⁶ and appear to represent intrinsically mediated class effects of FXR ligands, resulting from induction of pruritogenic cytokines such as IL-31¹⁴² and repression of BA synthesis with subsequent cholesterol buildup and reduced expression of the LDL-receptor,¹⁴³ respectively. Further studies will have to critically evaluate whether non-steroidal FXR ligands are associated with similar safety concerns as shown for OCA and whether the risks observed with OCA could also represent a class effect.

In addition to FXR, therapeutic targeting of its downstream target FGF-19 may also represent an interesting therapeutic strategy. Accordingly, engineered non-tumorigenic FGF-19 mimetics reduced hepatic inflammation and fibrosis in the Mdr2 KO mouse model of sclerosing cholangitis.^144,145 In people with PBC inadequately responding to UDCA, the FGF19 mimetic aldafermin improved cholestatic liver enzymes¹⁴⁶ while in PSC it improved non-invasive serum markers of hepatic fibrosis but lacked effects on cholestatic liver enzymes.¹⁴⁷ Notably, aldafermin led to a more hydrophilic, less toxic BA profile which correlated with changes of fibrosis markers and could explain its antifibrotic effects.¹⁴⁸

Given the role of FXR in modulating hepatic glucose and lipid metabolism and a perhaps underestimated role of cholestasis and altered BA metabolism in MASLD,^149,150 there may be a role for BA/FXR-targeted strategies, in particular as part of combination strategies including incretins/GLP-1R agonists (which are also stimulated by BA/TGR5). OCA was associated with an improvement of fibrosis without resolution of MASH,¹⁵¹ but has not been approved for this indication due to the unfavourable risk-benefit ratio, including concerns regarding dyslipidaemia. Similar to the area of cholestatic liver diseases, ongoing studies with non-steroidal FXR agonists have been promising but remain to demonstrate efficacy on prognostically relevant endpoints in combination with an appropriate safety profile.¹⁵²

BA therapies in gallstone disease: history and prospects

In the 1960s and early 1970s, Schoenfield and Thistle at the Mayo Clinic tested the effect of different BAs on cholesterol saturation of bile and found that CDCA reduced cholesterol saturation.¹⁵³ A first randomised trial comparing CDCA with CA administration demonstrated that CDCA led to gallstone dissolution in the majority of patients, which was not observed in the CA group.¹⁵⁴ The efficacy of 12-15 mg/kg CDCA on gallstone dissolution was about 50-70%.¹⁵⁵ Common side effects of CDCA treatment were hypertransaminasaemia and diarrhoea.¹⁵⁶ Conversion of CDCA to lithocholic acid was also a concern, particularly if not sulfated.¹⁵⁷ Subsequently, different randomised studies comparing CDCA with UDCA showed a similar efficacy of both BAs for gallstone dissolution but with less side effects for UDCA.^156,158 The combination of extracorporeal shock wave lithotripsy¹⁵⁹ with UDCA treatment enhanced the velocity of gallstone dissolution, but did not avoid disadvantages like long-term medical treatment, slow velocity of gallstone fragment dissolution (<1 mm decrease in size per month) and high rates of gallstone recurrence after cessation of medical treatment. With the upcoming and then widespread availability of laparoscopic cholecystectomy, the medical dissolution of gallstones with BA treatment lost its clinical importance. Currently, UDCA treatment is used for primary prevention of gallstone formation and secondary prevention of progression of asymptomatic gallstones to symptomatic disease. Administration of UDCA is recommended (e.g. 500 mg daily over 6 months) in people undergoing bariatric surgery as it reduces the risk of de novo gallstone formation as well as the development of symptomatic gallstone disease.^160-163 UDCA is also recommended as life-long treatment in people with LPAC syndrome for prevention of recurrent gallstone formation and associated complications.^164-166

BA therapies in genetic disorders of the child/adolescent

Cholestatic disorders comprise a substantial component of serious liver diseases in childhood. In brief, BA-targeted therapies have been successfully applied, and approved, for use in three different categories of genetic disorders: bile acid synthesis defects (BASDs) such as 3β-hydroxy-Δ5-C27-steroid oxidoreductase deficiency, Δ4-3-oxosteroid 5β-reductase deficiency, or cerebrotendinous xantomatosis; Alagille syndrome (ALGS); and monogenic disorders grouped together with the wide-ranging label of progressive familial intrahepatic cholestasis (PFIC).^166-168 The essential targets of these therapies are effective and distinct.

The treatment for BASDs is the end-product of BA synthesis, CA,^169,170 although CDCA treatment has also been reported to be effective.¹⁷¹ Deficiencies in key BA synthesis genes that comprise the BASDs lead to the unregulated production and intracellular accumulation of toxic BA intermediates. CA, as an FXR ligand, mediates feedback inhibition of the transcription of several key BA synthetic enzyme genes, strongly suppressing the aberrant production of these damaging molecular intermediates. In addition, CA enriches the composition of the BA pool with primary BAs, which improves nutritional deficiencies and accelerates growth in young patients.¹⁷²

In distinction to BASDs where the root pathogenic issue is inadequate primary BA synthesis, recently approved effective treatments for ALGS and PFICs target the intrahepatic accretion of BAs by interfering with ileal reclamation of BAs. Gutrestricted luminal inhibition of IBAT reduces the BA pool size and re-presentation of BAs to an already cholestatic liver.¹⁷³ These two BA-based therapies were developed based on a clearer understanding of the molecular bases of BA recirculation and synthesis.^174,175

Insufficiency of Jagged1 (JAG1) or its cell surface receptor NOTCH2 during key periods of gestation leads to the varying multi-organ manifestations of ALGS.^176,177 In patients with a liver phenotype, aberrant JAG1-NOTCH2 signalling leads to significantly impaired development and maintenance of the biliary tree, both intra- and extrahepatic. In the more severe forms of biliary maldevelopment, there is profound, early cholestasis that manifests clinically with conjugated hyperbilirubinemia, nutritional consequences of cholestasis and profound pruritus starting in infancy. The pruritus in ALGS is so debilitating that it often becomes the primary indication for LT when there is no relief. A major improvement in the availability of effective treatment options for pruritus in ALGS came from clinical studies of IBAT inhibitors in children.^174,178 In particular, with a novel study design, ICONIC demonstrated a marked improvement in pruritus on maralixibat,¹⁷⁹ while the ASSERT trial provided similar results with odevixibat.¹⁸⁰ Both agents have been approved by the EMA and FDA since 2021, and are currently available for use in children with ALGS as young as 3 months of age. IBAT inhibitors are generally well tolerated, although not unexpectedly, some patients will develop BA-induced diarrhoea that in children usually responds to dose reduction. Importantly, in addition to improvements in pruritus, clinical outcomes like growth, sleep and native liver survival appear to improve in patients with ALGS on maralixibat.^181,182

First explored in children with ATP8B1 (PFIC1) and ABCB11 (PFIC2) deficiencies suffering from marked pruritus, odevixibat was shown in the PEDFIC study to significantly reduce pruritus.¹⁸³ Maralixibat ameliorated pruritus in a group of patients with a variety of causes of PFIC (mostly PFIC2, but also ABCB4, TJP2, MYO5B and ATP8B1 deficiencies) in the MARCH-PFIC study.¹⁸⁴ IBAT inhibitors have been approved for use in patients with PFIC and significant pruritus since 2021. The more long-term clinical and symptomatic outcomes of IBAT inhibitors in patients with PFIC are beginning to be studied, but are expected to be similar to the improvements seen in patients with ALGS, as is the role of serum BA (sBA) monitoring. It is important to note that for both patients with PFIC and ALGS on IBAT inhibitors, upwards of 40-50% respond insufficiently, highlighting the need to explore combinatorial approaches in the near future, including with agents currently under study in adult cholestatic disorders, like NCA and FXR or PPAR agonists.

The efficacy of other medications to treat paediatric genetic cholestatic disorders remains to be systematically studied, but heterozygous ABCB4-deficient paediatric patients appear to respond to UDCA like their adult counterparts.¹⁸⁵ Roles for medications like rifampicin, bezafibrate and a variety of other anti-pruritic agents have not been systematically studied in affected children, and it is up to the clinician to decide their roles in the current era.

BA therapies in genetic disorders of the adult

While prospective randomised trials on the use of BA therapies (UDCA, FXR agonists, IBAT inhibitors, NTCP inhibitors) are missing for adults with genetic intrahepatic cholestasis, UDCA has been widely used in these patients and is recommended as first-line therapy in patients with (heterozygous) ABCB4 variants.¹⁶⁶ Variants in the ABCB4 gene, which encodes the aminophospholipid floppase MDR3 underlie a broad spectrum of clinical phenotypes, including isolated elevation of gamma-glutamyltransferase, ICP, LPAC, biliary fibrosis/cirrhosis and CCA.^185-188 UDCA at doses ranging from 8-10 mg/kg of body weight up to 13-15 mg/kg of body weight is recommended life-long for people with LPAC syndrome, based on an initial description of six patients who showed complete resolution of gallstone-associated symptoms and normalisation of liver enzymes.^164-166 In a retrospective cohort study of 67 adult carriers of heterozygous potentially pathogenic ABCB4 variants of whom 93% received UDCA treatment, LT-free survival was 91% over a median follow-up of 14 years.¹⁸⁸ The efficacy of UDCA to improve or normalise serum liver tests in patients with ABCB4 disease, carrying at least one missense variant, has been shown mainly in paediatric but also in adult cohorts.^185,189,190 Based on these observations, UDCA is recommended for people carrying at least one ABCB4 missense variant.¹⁶⁶ Whether UDCA can improve native liver survival and reduce the risk of CCA development in ABCB4 disease remains to be determined.

As UDCA and its taurine conjugate TUDCA have been shown to increase the amount of functional BSEP in the canaliculular membrane of hepatocytes, UDCA has been widely used in various subtypes of genetic intrahepatic cholestasis including mono- and biallelic BSEP deficiency^58,191-195 and DILI.¹⁹⁶ A recent systematic review found that UDCA administration improved serum liver tests in DILI; however, the lack of prospective randomised, controlled trials prevents a clear recommendation for UDCA in DILI at present.¹⁹⁶

Inhibition of BA uptake from the small intestine using IBAT inhibitors has been tested in several phase II and III trials in paediatric patients with ALGS and PFIC.^{179,180,183,184} However, data in adult patients, especially those with monoallelic variants in a PFIC-associated gene, is scarce at the moment.¹⁹⁷ Further studies are needed to elucidate which adult patients benefit from IBAT inhibitor therapy and whether the treatment needs to be administered continuously or intermittently in episodic cholestasis.

BA therapies in pregnancy

Normal gestation results in relative hypercholanaemia due to reproductive hormone-mediated reduced FXR function and competitive inhibition of biliary transporters.^198-201 Thus, the normal range for sBA concentrations is higher in pregnancy than in non-pregnant individuals and a non-fasting threshold of ≥19 μmol/L would improve diagnostic accuracy for ICP vs. normal pregnancy.²⁰² ICP is the most common pregnancy-specific liver disorder, affecting 0.7% of pregnant women of European ancestry and approximately twice as many South Asian women,²⁰³ and 3-4 times as many women from Andean nations.²⁰⁴ The diagnosis of ICP is confirmed if pruritus and abnormal liver tests resolve postpartum.²⁰⁵ Women with non-gestational cholestatic disorders may have an exacerbation of pruritus or hypercholanaemia in pregnancy, and some present de novo in pregnancy.^206-208 Thorough investigation usually enables this group of women to be distinguished from those with ICP.

Women with ICP have increased rates of adverse pregnancy outcomes which are related to the maximal maternal sBA concentration.^209-211 Specifically, those with non-fasting sBA ≥100 μmol/L have increased risk of stillbirth, and those with postprandial sBA ≥40 μmol/L have increased risk of spontaneous preterm birth, foetal distress and admission to the neonatal unit.²¹¹ While there is limited literature to evaluate the implication of hypercholanaemia for pregnant women with pre-gestational cholestasis, existing studies of women with PBC and PSC indicate that the same thresholds for sBA concentrations are likely to predict pregnancies at risk of these adverse outcomes.²¹² Unlike in non-pregnant individuals, when measuring sBA concentrations in pregnancy it is preferred to assay non-fasting blood samples, as this gives an indication of the peak sBA concentration, enabling identification of pregnancies at risk of stillbirth and spontaneous preterm birth.²⁰²

Interventions to reduce sBA concentrations, or to modify the hydrophobicity of bile, are likely to reduce the risk of adverse pregnancy outcomes in ICP. While the precise aetiology of ICP-associated stillbirth is not understood, it does not appear to relate to foetal growth restriction or placental insufficiency. Ex vivo placental studies have shown BA-induced placental vasoconstriction that may exacerbate foetal risk.²¹³ Human observational studies using foetal electrocardiogram monitors have demonstrated BA-induced foetal arrhythmia in ICP pregnancies, and that UDCA treatment can partially normalise the abnormalities in foetal cardiac function and rhythm that are associated with hypercholanaemia.²¹⁴ UDCA has also been shown to reduce the risk of spontaneous preterm birth in singleton pregnancies in an individual participant meta-analysis that included data from 6,974 ICP pregnancies;²¹⁵ additionally, it reduced a variety of adverse pregnancy outcomes in an aggregate meta-analysis when compared to other treatments or placebo.²¹⁶ There is less evidence for alternative therapies, although small studies indicate that S-adenosyl methionine, cholestyramine, bezafibrate and guar gum may improve the severity of pruritus and biochemical derangements in some cases.²⁰⁵ A large, multinational trial is currently comparing the impact of rifampicin and UDCA on pruritus in ICP.²¹⁷ There are limited data on the role of IBAT inhibitors in ICP; the OHANA trial that aimed to evaluate volixibat only recruited four participants and was stopped; three had improvements in pruritus severity and sBA concentrations, but diarrhoea was a side effect when higher doses were used.²¹⁸ Approximately 30% of ICP cases have pathogenic mutations in biliary transporters (ABCB11, ABCB4, ATP8B1),^219,220 and others have common variants that confer increased susceptibility to cholestasis.²²¹ It is likely that improved understanding of the genetic background of individual women with ICP may enable better prediction of which cases will respond to specific treatments. At present recommended treatment is to give UDCA in women with sBA ≥40 μmol/L to protect against spontaneous preterm birth, and possibly also stillbirth, and to perform delivery at 35-36 weeks’ gestation if sBA levels are ≥100 μmol/L to reduce the risk of stillbirth.²⁰⁵ It is hoped that ongoing studies will clarify the role of rifampicin and possibly IBAT inhibitors.

For women with ICP, it is important to consider the impact of the diagnosis on future health. Affected women²²² and their first degree relatives²²³ have increased risk of hepatobiliary disorders, including hepatic fibrosis and malignancy. Thus, for those with severe disease or a family history of hepatobiliary disorders a cholestasis gene panel may identify a subgroup of cases that should be referred for ongoing hepatology follow-up. Similarly, investigation of those whose symptoms do not resolve postpartum will identify women at risk of future liver pathology. It is likely that improved understanding of the underlying aetiology of hypercholanaemia in pregnancy will identify women (and pedigrees) that will benefit from longer term BA therapies to maintain long-term health.

Prospects for BA therapies

UDCA represents the best-established therapeutic BA for the treatment of cholestatic liver diseases such as PBC, most likely PSC, ICP, and ABCB4 deficiency (including LPAC syndrome), and for the prevention of gallstone formation under conditions of rapid weight loss (Table 1). Awareness has increased in relation to adequate dosing, particularly in PSC, to avoid excessive formation of potentially toxic BA metabolites like LCA and its metabolites in the intestine and hepatobiliary tract. Additional indications for UDCA within and outside the liver are under discussion (Table 1) but not established. Combination treatments of UDCA, a post-translational secretagogue, with transcriptional modulators, i.e. nuclear receptor agonists for FXR, PPAR or PXR, are promising and in part approved for treatment of fibrosing cholangiopathies. UDCA hybrids for treatment of hepatobiliary malignancies are under investigation.

Whether BA uptake inhibitors of IBAT/ASBT at the level of the ileum or of NTCP at the level of the hepatocyte (Fig. 5²²⁴) have an additive beneficial long-term effect in chronic cholestatic liver diseases of the adult remains to be determined; notably, they also impair uptake of beneficial UDCA conjugates.

Fig. 5. The enterohepatic circulation of BAs.

BSEP (ABCB11) is the main canalicular export system for BAs in humans. After reabsorption by ASBT (IBAT/SLC10A2) in the ileum, BAs are transported back to the liver through the portal circulation. Hepatic reuptake of conjugated BAs from portal blood is mainly mediated by NTCP (SLC10A1), thus completing the enterohepatic circulation. A small fraction of BAs escape hepatic clearance and enter the systemic circulation. ASBT is also expressed in the proximal tubules of the kidney. In the terminal ileum, BAs activate the nuclear receptor FXR, which induces FGF19. FXR-activated FGF19 binds to the FGFR4/β-Klotho complex in the liver, reducing BA synthesis by repressing CYP7A1. FXR also induces transcriptional repressors that directly inhibit CYP7A1 and NTCP expression in the liver while stimulating BSEP expression (not shown). Colonic TGR5-mediated GLP-1 secretion may exert both hepatoprotective and cholangioprotective effects. BAs also interact with gut microbiota, which metabolise primary BAs into secondary BAs (Figure modified from Trauner et al.²²⁴). ASBT/IBAT, apical sodium-dependent bile acid transporter; BAs, bile acids; BSEP, bile salt export pump; CYP7A1, cholesterol 7 alpha-hydroxylase; FGF19, fibroblast growth factor 19; FGFR4, fibroblast growth factor receptor 4; FXR, farnesoid X receptor; GLP-1, glucagon-like peptide 1; NTCP, Na⁺-taurocholate cotransporting polypeptide; TGR5, Takeda G-protein–coupled receptor 5.

NCA is a promising BA for cholangiopathies, possibly as an add-on to UDCA therapy, but phase III data, mainly in PSC, will be needed prior to further judgement, and a role for NCA in other hepatobiliary diseases, e.g. PBC, requires evaluation.

Novel BA therapies could include polyhydroxylated BAs (PHBAs) spontaneously formed in Bsep/Abcb11 deficient mice, which protect these mice from the severe liver injury seen in human PFIC2.^225,226 However, challenging Bsep KO mice with exogenous BAs such as CA induces a more pronounced PFIC2-like phenotype.²²⁷ Intriguingly, Bsep KO mice are protected from acquired cholestatic liver injury,²²⁸ which may be explained by metabolic preconditioning resulting in the formation of more hydrophilic, less toxic, PHBAs which also inhibit Egr1-mediated hepatic inflammation.²²⁹ As such deletion of Bsep in Mdr2 KO mice or treatment with PHBAs also attenuates sclerosing cholangitis.^229,230 Notably individuals with PFIC2 with a more benign clinical course show higher intrinsic levels of PHBAs.²³¹ Finally, PXR and CAR agonists profoundly stimulated BA-hydroxylating/detoxifying enzymes Cyp3a11 and Cyp2b10 resulting in formation of PHBAs and preventing cholestatic liver injury in mice,²³² pointing towards an anticholestatic mechanism for such compounds (e.g. rifampicin, phenobarbital)²³² which are mainly used for symptomatic treatment of pruritus and jaundice, but also could have more direct anticholestatic potential, as exemplified by the effects of rifampicin in the treatment of severe persistent hepatocellular secretory failure.^233,234

Finally, unusual microbial metabolites such as oxo-, iso- and allo-BAs may have anti-inflammatory and immunomodulatory effects on T cells by acting on BA-activated nuclear receptors, such as FXR and VDR, as recently reviewed,¹²¹ further pointing towards the still untapped potential of using novel BA derivatives as immunomodulatory drugs. There has been an impressive surge in the identification of novel BA metabolites, including previously unknown amino acid and polyamine conjugates, some of which may exert FXR-mediated or other, as yet unidentified, signalling functions.^121,235-237 Future studies should explore the plethora of novel microbial BA metabolites which may be encountered in the small intestine, in addition to the large intestine and faeces.²³⁸

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jhep.2025.06.010.

Keypoints.

• Bile was first proposed as a potential remedy more than 3,500 years ago and bear bile as ‘king of animal biles’ has been used for the treatment of jaundiced individuals for more than two millennia in China.

• Ursodeoxycholic acid (UDCA) is a dominant bile acid of Chinese black bears, but forms only 1-3% of the human bile acid pool and is a potent post-translational secretagogue in hepatocytes and cholangiocytes, exerting anticholestatic, anti-apoptotic, anti-inflammatory and antifibrotic effects in cholestatic liver diseases.

• UDCA is the evidence-based first-line medical treatment for primary biliary cholangitis (PBC) and is also used for other cholestatic disorders, including intrahepatic cholestasis of pregnancy, albeit the supporting evidence is weaker.

• UDCA’s litholytic activity in cholesterol gallstone disease has lost relevance in daily practice.

• Norucholic acid (NCA) is a side chain-shortened derivative of UDCA with promising anticholestatic effects in fibrosing cholangiopathies, mediated through molecular signalling mechanisms distinct from those of UDCA.

• The potent steroidal farnesoid X receptor (FXR) agonist obeticholic acid (OCA) exerts additive anticholestatic and antifibrotic effects in individuals with PBC with incomplete responses to UDCA, but might induce hepatotoxicity in decompensated cirrhosis when overdosed and was denied full approval by American and European regulators.

• Novel UDCA derivatives such as Bamet-UD2, a UDCA-cisplatin hybrid, or UDCA-histone deacetylase 6 inhibitors are being tested in gastrointestinal oncology and polycystic liver diseases with promising initial results.

• Bile acid therapies in rare paediatric genetic bile acid synthesis disorders improve prognosis.

• Modulation of the enterohepatic circulation of endogenous bile acids with ileal bile acid transporter inhibitors has shown promise in rare paediatric cholestatic disorders and is being tested in adults.

Abbreviations

ASBT

apical sodium-dependent bile acid transporter

BA

bile acid

BASD

bile acid synthesis defects

BSEP

bile salt export pump

CA

cholic acid

CDCA

chenodeoxycholic acid

DCA

deoxycholic acid

DILI

drug-induced liver injury

FXR

farnesoid X receptor

IBAT

ileal bile acid transporter

IBD

inflammatory bowel disease

ICP

intrahepatic cholestasis of pregnancy

LCA

lithocholic acid

LPAC

low phospholipid associated cholelithiasis

LT

liver transplant

NCA

norucholic acid

OCA

obeticholic acid

PBC

primary biliary cholangitis

PFIC

progressive familial intrahepatic cholestasis

PHBAs

polyhydroxylated Bas

PKCα

protein kinase Cα

PLD

polycystic liver disease

PPAR

peroxisome proliferator-activated receptor

PSC

primary sclerosing cholangitis

TUDCA

tauroursodeoxycholic acid

UDCA

ursodeoxycholic acid