EASL–AASLD PBC Consensus 2026: Journal of Hepatology | September 2026 | EASL–AASLD Delphi Consensus
- ALP + bilirubin remain the core biochemical treatment targets
Both ALP and total bilirubin at 12 months consistently predict clinically important outcomes in PBC, including decompensation, transplantation and death.
- Aim for complete biochemical normalisation when possible
For phase III trials, normal ALP + normal bilirubin is considered the ideal response; ALP <1.5–1.67× ULN with normal bilirubin remains an acceptable response. Complete normalisation carries stronger prognostic information.
- Loss of biochemical response matters—even later
An inadequate biochemical response after 6–12 months of UDCA, or at any subsequent time, identifies patients at increased risk of poor outcomes.
- AST, ALT and GGT are not established long-term surrogate markers
Unlike ALP and bilirubin, current evidence is insufficient to use AST, ALT or GGT as validated predictors of long-term treatment benefit.
- Albumin and platelets add important prognostic information
Both albumin and platelet count during treatment are associated with clinical outcomes and should not be overlooked when assessing disease trajectory.
- FibroScan adds prognostic information beyond biochemistry
VCTE liver stiffness measurement (LSM) predicts clinical outcomes independently of biochemical markers. An LSM >10 kPa is associated with worse outcomes.
- Change in liver stiffness may be as important as the absolute value
A >20% increase in LSM over 1 year has been associated with worse outcomes, whereas a >20% decrease is associated with better outcomes. However, prospective validation remains necessary.
- Future PBC assessment should combine biochemistry + fibrosis
The consensus strongly supports a composite endpoint combining biochemical response with stable or decreased LSM, particularly in studies lasting at least 24 months in early-stage disease.
- Real-world evidence may become central to PBC drug development
Because conventional long-term placebo-controlled trials are increasingly difficult and sometimes ethically problematic, high-quality real-world evidence (RWE) can complement randomized trials and potentially support regulatory decisions.
- External real-world control arms are acceptable when RCTs are impractical
The panel supports rigorously designed single-arm trials with RWD-derived external controls, including target-trial emulation, when conventional randomised trials are not feasible.
- Real-world evidence must be “regulatory grade”
Registries, EHRs, claims databases and distributed networks can all contribute, but data must be fit-for-use, endpoints harmonised, and biases prespecified and actively addressed.
- Symptoms are independent therapeutic targets
A patient can have an excellent biochemical response and still have substantial pruritus, fatigue, cognitive symptoms and impaired quality of life. These symptoms warrant assessment and treatment independent of disease-progression endpoints.
- PBC-40 should be the core disease-specific quality-of-life tool
The consensus recommends PBC-40 as the primary disease-specific instrument; generic tools such as EQ-5D, SF-36 and PROMIS-29 should not be used alone to assess PBC quality of life.
- Pruritus: WI-NRS is the preferred tool—and 3 points matter
The Worst Itch Numerical Rating Scale (WI-NRS) is recommended for pruritus assessment, with a ≥3-point reduction representing the primary threshold for clinically meaningful improvement.
- Fatigue must be measured separately from liver disease severity
Fatigue affects approximately 40–80% of patients and may persist despite biochemical response. The PBC-40 fatigue domain is recommended as the primary disease-specific measure, complemented by PROMIS Fatigue SF-7a; wearables remain exploratory.
GastroAGI Bottom Line
The major conceptual shift is from “ALP response” to multidimensional PBC control. Future assessment should increasingly integrate biochemical response + fibrosis trajectory + patient-reported symptoms. ALP and bilirubin remain central, but LSM adds independent prognostic information, while pruritus, fatigue and quality of life must be treated as genuine outcomes rather than secondary concerns. The paper also establishes a roadmap for using rigorous real-world evidence when traditional confirmatory trials are impractical.
