What is the role of prognostic scores, such as the Mayo Risk Score or the UK-PBC (United Kingdom-Primary Biliary Cholangitis) score, in managing a middle-aged woman with Primary Biliary Cholangitis (PBC) and potential disease progression?

Prognostic Scores in Primary Biliary Cholangitis

In PBC patients on UDCA therapy, use the GLOBE score and UK-PBC score as continuous quantitative risk stratification tools after 12 months of treatment, as both demonstrate superior discriminatory performance compared to qualitative criteria and traditional scores like MELD or Child-Pugh. 1

Baseline Risk Stratification (Before or At Treatment Initiation)

At diagnosis, distinguish early from advanced disease stage using: 1

• Liver stiffness measurement (LSM) by transient elastography: LSM <10 kPa indicates early disease; LSM ≥10 kPa indicates advanced disease
• Biochemical parameters: Normal albumin AND bilirubin = early stage; abnormal albumin OR bilirubin = advanced stage
• Histology when available: Absent/mild fibrosis vs. bridging fibrosis/cirrhosis

On-Treatment Risk Stratification (After 12 Months of UDCA)

Primary Quantitative Scores (Preferred)

GLOBE Score 1

• Components: Age, total bilirubin, alkaline phosphatase (ALP), albumin, platelet count
• Performance: Accurately predicts liver transplant-free survival at 5 and 10 years (c-statistics 0.81-0.82)
• Validation: Externally validated in multiple international cohorts; superior to qualitative criteria 2

UK-PBC Score 1, 3

• Components: Baseline albumin and platelet count, plus bilirubin, AST or ALT, and ALP at 12 months after starting UDCA
• Performance: Predicts risk of major outcomes (liver-related death, liver transplant, or bilirubin >100 µmol/L) at 5,10, and 15 years with AUROCs of 0.96,0.95, and 0.94 respectively 1
• Validation: Externally validated; superior to qualitative criteria, MELD, and Child-Pugh scores 1

Comparative Performance: In a large international cohort of 1,100 patients, the GLOBE score showed slightly superior discriminatory performance (concordance statistic 0.80 at 1 year) compared to UK-PBC (0.74) and Mayo Risk Score (0.76), though differences were not statistically significant 2

Qualitative Response Criteria (Alternative)

Use Paris-I, Paris-II, Rotterdam, Toronto, Rochester, Barcelona, or Ehime criteria to define biochemical response 1

Additional Risk Refinement Tools

APRI Score 1

• APRI >0.54 after 12 months of UDCA therapy independently predicts risk of cirrhosis decompensation
• Combine with GLOBE score biochemical non-response for improved risk stratification

Gamma-Glutamyl Transferase (GGT) 1

• GGT >3.2-fold the upper limit of normal at 12 months identifies patients at increased risk of liver transplant or liver-related death, independent of ALP values

Optimal Treatment Response Markers 1

• ALP normalization OR serum bilirubin <0.6× upper limit of normal after 12 months = lowest risk for liver transplant or death

ELF Score 1

• Associated with clinical outcomes in PBC
• Can be used as an additional risk stratification tool

Longitudinal Monitoring with Liver Stiffness

Serial LSM by Transient Elastography 1

• Worsening LSM predicts patient outcomes during follow-up
• An increase of 2.1 kPa/year in LSM associates with an 8.4-fold increase in risk of adverse outcomes
• Recommended frequency:
  • Every 2 years in patients with early-stage disease
  • Every year in patients with advanced-stage disease

Practical Algorithm for Clinical Use

1. At baseline: Assess disease stage using LSM (<10 vs. ≥10 kPa), albumin, bilirubin, and platelets 1

2. After 12 months of UDCA: Calculate GLOBE and/or UK-PBC scores 1

3. Refine risk assessment: Add APRI score and GGT levels to identify high-risk patients despite apparent biochemical response 1

4. Monitor longitudinally: Repeat LSM at appropriate intervals based on disease stage 1

5. Identify optimal responders: Target ALP normalization or bilirubin <0.6× ULN for best outcomes 1

Critical Pitfalls to Avoid

Do not rely on MELD or Child-Pugh scores for early-stage PBC risk stratification, as they were developed for cirrhotic patients and perform poorly in early disease 1, 2

Do not use qualitative response criteria alone when quantitative scores (GLOBE, UK-PBC) are available, as the latter demonstrate superior discriminatory performance 1

Do not ignore patients who achieve biochemical response by qualitative criteria if they have elevated APRI (>0.54) or GGT (>3.2× ULN), as these indicate residual risk 1

Be cautious interpreting LSM in patients with elevated bilirubin due to extrahepatic bile duct obstruction, as this can overestimate fibrosis stage 1

Evidence Quality Considerations

The 2021 EASL guidelines provide the most recent and comprehensive recommendations, giving strong recommendations (Level of Evidence 3) for using continuous risk scores (GLOBE and UK-PBC) during treatment 1. Recent research from 2024 and 2023 confirms these scores maintain excellent predictive performance in real-world cohorts, including those with advanced disease 4, 5. The Mayo Risk Score, while historically important and still useful (particularly in advanced disease 5), has been superseded by the more accurate GLOBE and UK-PBC scores for most clinical applications 1, 2.