Second-Line Treatment Options After Inadequate UDCA Response in PBC

Key Takeaways:

• In the United States, the available second-line options after an inadequate response to or intolerance of ursodeoxycholic acid (UDCA) are the peroxisome proliferator-activated receptor (PPAR) agonists – elafibranor and seladelpar – with off-label fibrates as an additional PPAR-targeted choice. The farnesoid X receptor (FXR) agonist obeticholic acid, the original second-line agent, was withdrawn from the US market in late 2025.
• Each agent was studied against a biochemical-response endpoint anchored on alkaline phosphatase (ALP) and bilirubin; selection now weighs the PPAR options against one another and against fibrates, considering fibrosis stage, pruritus burden, and tolerability.
• Second-line therapy is a longitudinal decision: Response is reassessed on the same ALP and bilirubin framework, and the agent can be reconsidered if response remains inadequate.

When a patient with primary biliary cholangitis (PBC) has an inadequate response to – or cannot tolerate – first-line UDCA, the question is no longer whether to escalate but which agent to add. The second-line landscape shifted substantially in 2024 and 2025, and framing the options by mechanism rather than by individual product helps match therapy to the patient in front of you.¹

The PPAR agonists now available

PPAR agonists are the mainstay of second-line PBC therapy in the United States. Elafibranor, a dual PPAR-alpha/delta agonist, received accelerated approval in June 2024. In the phase 3 ELATIVE trial, it achieved a biochemical response in 51 percent of patients versus 4 percent with placebo at week 52, with ALP normalization in 15 percent.^2,3

Seladelpar, a selective PPAR-delta agonist, received accelerated approval in August 2024; in the phase 3 RESPONSE trial, it produced a biochemical response in 62 percent versus 20 percent at 12 months, with ALP normalization in 25 percent.^4,5

Both approvals rest on ALP reduction, a surrogate endpoint, and a clinical-outcome benefit has not yet been demonstrated.^3,5

The pan-PPAR agonist bezafibrate, used off-label, produced a complete biochemical response in 31 percent versus 0 percent over 24 months in the BEZURSO trial.⁶

The withdrawn FXR agonist

Obeticholic acid, an FXR agonist, was the first second-line agent approved for PBC, and the phase 3 POISE trial established its biochemical effect, with the composite endpoint met in 46 to 47 percent of patients versus 10 percent with placebo.³

Its labeling had restricted use to patients without cirrhosis or with compensated cirrhosis and no portal hypertension, and dose-dependent worsening of pruritus was its most common reason for discontinuation.⁷

Obeticholic acid was withdrawn from the US market in late 2025; however, Intercept voluntarily removed it effective November 14, 2025, at the FDA's request, and the FDA withdrew approval effective November 24, 2025, after the confirmatory trial failed to verify clinical benefit and showed serious liver injury, including in patients with early-stage disease. The FXR agonist class therefore currently has no marketed option for PBC in the United States; its trial evidence remains useful only as class context.⁸

Matching the agent to the patient

Because the available agents were not compared head-to-head, selection among the PPAR options is individualized. Pruritus is one axis: PPAR agonists have shown neutral-to-favorable effects on itch, with seladelpar significantly reducing pruritus in its trial while elafibranor did not separate from placebo on its itch endpoint.^2,4

Tolerability is another – fibrates can raise serum creatinine and cause myalgia, and the PPAR agonists warrant attention to muscle symptoms, particularly alongside a statin.^2,7

Lipid profile and fibrosis stage also inform the choice, and selection ultimately rests on individual patient priorities, symptom burden, and comorbidities.¹

Whichever agent is chosen, response is reassessed on the same ALP and bilirubin framework used to identify the inadequate first-line responder, and therapy can be changed if the target is not met.^9,10