Obeticholic acid in biliary atresia: a pediatric FXR trial stopped by an adult safety signal

What was tried

Obeticholic acid is a semisynthetic bile acid and a potent agonist of the farnesoid X receptor (FXR, gene NR1H4, ENSG00000012504), with ChEMBL identifier CHEMBL566315 and a maximum phase of 4. NCT06121375 was a randomized, double-blind, placebo-controlled Phase 2/3 study run by Intercept Pharmaceuticals to assess efficacy, safety, pharmacokinetics, and pharmacodynamics of obeticholic acid in pediatric patients with biliary atresia (ClinicalTrials.gov NCT06121375). The design compared oral obeticholic acid against matching placebo, with primary endpoints of composite liver-related clinical events and changes in the Pediatric End-stage Liver Disease and MELD-Na scores. The study enrolled 28 participants. Intercept terminated it with the posted reason that the company made a business decision to stop the study based on the FDA request for voluntary withdrawal of Ocaliva and the clinical hold placed on all US IND studies involving obeticholic acid.

The biological hypothesis

FXR is the master bile acid sensor. Activation represses CYP7A1 and bile acid synthesis, induces FGF19 signaling, and promotes bile acid efflux, which reduces intrahepatic bile acid load in cholestatic disease. Biliary atresia is a progressive fibro-obliterative cholangiopathy of infancy in which bile flow is obstructed and retained bile acids drive hepatocyte injury and fibrosis, so an FXR agonist that lowers bile acid burden and slows fibrosis is a defensible mechanistic bet. Open Targets scores the NR1H4 to biliary atresia association at 0.2194, supported mainly by animal model, literature, and clinical evidence with no direct genetic association for the pair (Open Targets Platform, 2025). The target has strong small-molecule tractability, including an approved drug and high-quality pockets and ligands. The hypothesis was that a pathway already validated in adult cholestasis could be extended to a pediatric obstructive cholangiopathy with high unmet need.

What actually happened

The trial did not fail on its own efficacy data. It was halted by a regulatory action in the parent indication. On September 11, 2025, Intercept announced the voluntary withdrawal of Ocaliva for primary biliary cholangitis from the US market at the request of the FDA, and the agency placed a clinical hold on all Intercept trials conducted under a US IND involving obeticholic acid (Intercept press release, 2025). The driver was serious liver injury, including cases in patients without cirrhosis, in the primary biliary cholangitis population. The pharmacovigilance record aligns with a hepatobiliary and pruritus burden. Across 6798 OCALIVA reports in OpenFDA FAERS, the most frequent terms were pruritus at 2365 reports, or 34.8 percent, fatigue at 881, and death at 369, or 5.4 percent, with jaundice, cholelithiasis, and acute cholecystitis also present (OpenFDA FAERS, retrieved 2026-07-08). The biliary atresia study stopped at 28 enrolled before any efficacy conclusion could be drawn.

Failure mechanism, best guess

The proximate failure archetype is a safety signal, transmitted by read-across rather than generated inside the pediatric trial. The mechanistic tension is dose and disease context. FXR agonism lowers bile acid synthesis, but obeticholic acid is itself a bile acid analog, and in patients with compromised hepatic reserve the added bile acid load and the on-target cholestatic pruritus can worsen rather than protect. In advanced or cirrhotic livers the therapeutic window narrows, which is what surfaced in primary biliary cholangitis and what prompted the withdrawal. Biliary atresia infants have severe, evolving cholestasis and limited hepatic reserve, so the same on-target liability that closed the adult indication would be expected to apply, arguably with less margin. The clinical hold made the pediatric risk-benefit untenable regardless of any early pediatric signal.

How to prevent this next time

Two quantitative tools would have improved the position. First, a formal historical base-rate adjustment and competitive red-team analysis before committing a pediatric Phase 2/3. The adult FXR safety record, including dose-dependent hepatotoxicity and a boxed warning history in primary biliary cholangitis, should have entered the prior as a defined hazard, lowering the expected value of extension into a more fragile population. Second, a Bayesian safety-first stopping framework that treats serious liver injury as the gating endpoint, evaluating the predictive probability of an acceptable benefit-risk at each interim:

PP(success)=∫θ​P(final success∣θ,interim data)⋅π(θ∣interim data)dθ

With a prior informed by the adult hepatotoxicity rate and a hard safety threshold, the predictive probability of a favorable outcome would have stayed low until pediatric hepatic safety was demonstrated at low exposure, which argues for a dose-finding safety lead-in rather than a combined Phase 2/3. Historical base rates support the caution, since hepatology and rare pediatric programs carry approval probabilities well below the roughly 13.8 percent all-indication average from Phase 1 (Wong, Siah, and Lo, 2019). The single highest leverage change would have been ...

What this means for similar programs

FXR agonism remains an active mechanism, with cilofexor, tropifexor, and non-bile-acid agonists in development, and the central lesson is that on-target hepatotoxicity travels with the mechanism into new indications. Programs extending an adult cholestasis drug into pediatric obstructive cholangiopathy should not assume the adult safety envelope holds in infants with less hepatic reserve. A dose-finding safety lead-in with strict stopping rules, and a preference for agonists with wider windows, would reduce the chance that a parent-indication signal collapses a pediatric program mid-study.

Open questions

What exposures were reached in the 28 enrolled infants before the hold, and was there any early efficacy or safety signal. Would a lower dose or an intermittent schedule have preserved FXR pharmacodynamics while staying under the hepatotoxic threshold. Do non-bile-acid FXR agonists carry the same on-target liver injury risk in cholestatic pediatric disease. Will the biliary atresia hypothesis be retested with a different FXR agonist now that obeticholic acid is withdrawn.

Sources

1. • ClinicalTrials.gov, NCT06121375, full record, terminated status, whyStopped text, enrollment 28, Phase 2/3 design and endpoints: https://clinicaltrials.gov/study/NCT06121375 - Intercept Pharmaceuticals, voluntary withdrawal of OCALIVA for primary biliary cholangitis and clinical hold on obeticholic acid INDs, September 11, 2025: https://www.globenewswire.com/news-release/2025/09/11/3148535/23024/en/Intercept-Announces-Voluntary-Withdrawal-of-OCALIVA-for-Primary-Biliary-Cholangitis-PBC-from-the-US-Market-US-Clinical-Trials-Involving-Obeticholic-Acid-Placed-on-Clinical-Hold.html - Open Targets Platform, NR1H4 (ENSG00000012504) to biliary atresia (MONDO_0008867) association 0.2194, tractability: https://platform.opentargets.org/target/ENSG00000012504 - ChEMBL, obeticholic acid CHEMBL566315, small molecule, maximum phase 4: https://www.ebi.ac.uk/chembl/g/#search_results/all/query=obeticholic%20acid - OpenFDA FAERS, OCALIVA event counts, 6798 total reports, pruritus 2365, fatigue 881, death 369, retrieved 2026-07-08: https://api.fda.gov/drug/event.json?search=patient.drug.medicinalproduct:%22OCALIVA%22 - Farnesoid X Receptor Regulation of the NLRP3 Inflammasome Underlies Cholestasis-Associated Sepsis. Cell Metabolism.

2. - Discovery of farnesoid X receptor and its role in bile acid metabolism. Molecular and Cellular Endocrinology.

3. - Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019;20(2):273-286.