What is the role of Abic (Model for End-Stage Liver Disease) and Glasgow scores in assessing disease severity and guiding treatment for patients with alcoholic hepatitis?

ABIC and Glasgow Scores in Alcoholic Hepatitis

Both the ABIC score and Glasgow Alcoholic Hepatitis Score (GAHS) are validated prognostic tools that stratify mortality risk and guide treatment decisions in alcoholic hepatitis, with GAHS particularly useful for identifying patients who benefit from corticosteroid therapy (GAHS ≥9) and ABIC excelling at risk stratification into low, intermediate, and high mortality categories at 90 days and 1 year. 1

Understanding the Scoring Systems

ABIC Score Components and Interpretation

• Formula: [age (years × 0.1)] + [serum bilirubin (mg/dL) × 0.08] + [serum creatinine (mg/dL) × 0.3] + (INR × 0.8) 1
• Risk stratification using cutoff values of 6.71 and 9.0: 1
  • Low risk (ABIC <6.71): 0% mortality at 90 days, excellent prognosis without specific therapy 1, 2
  • Intermediate risk (ABIC 6.71-9.0): 30% mortality at 90 days, may benefit from corticosteroid or pentoxifylline therapy 1, 2
  • High risk (ABIC >9.0): 75% mortality at 90 days, requires aggressive treatment consideration 1, 2
• The ABIC score predicts both 90-day and 1-year mortality, making it useful for longer-term prognostication compared to other static models 1

Glasgow Alcoholic Hepatitis Score (GAHS) Components and Interpretation

• Scoring parameters (each assigned points): 1
  • Age: <50 years vs ≥50 years
  • White cell count: <15 vs ≥15 × 10⁹/L
  • Urea: <5 vs ≥5 mmol/L
  • PT ratio: <1.5 vs 1.5-2.0 vs >2.0
  • Bilirubin: <125 vs 125-250 vs >250 μmol/L
• Critical threshold GAHS ≥9: Predicts >50% mortality at 28-84 days and identifies patients who benefit from corticosteroid treatment 1, 2
• GAHS <9: Patients have favorable outcomes (5.9% 28-day mortality) with no survival benefit from corticosteroids 3, 4
• The GAHS on day 1 predicts 28-day survival with 81% accuracy, comparatively higher than the modified discriminant function 2, 3

Clinical Application Algorithm

Step 1: Initial Risk Stratification at Presentation

• Calculate both ABIC and GAHS scores at admission along with MELD and mDF scores 1, 5
• GAHS ≥9 identifies treatment candidates: These patients have significantly better 90-day outcomes with treatment (58% survival) versus no treatment (30% survival, p=0.01) 3
• ABIC score stratifies risk independently: Use to counsel patients on prognosis and intensity of monitoring 1
• Patients with consistently low scores (ABIC <6.71, GAHS <9, MELD <25) have low mortality (6-9% at 28 days) and do not require corticosteroid therapy 4, 6

Step 2: Treatment Decision Framework

• For GAHS ≥9 without contraindications: Initiate methylprednisolone 32 mg daily after excluding active infection 5, 3
• For ABIC intermediate risk (6.71-9.0): Consider corticosteroid or pentoxifylline therapy, as these patients have 30% 90-day mortality 1
• For ABIC high risk (>9.0) with MELD >26: Refer for liver transplantation evaluation given 75% 90-day mortality 1, 5
• Screen for infections (blood, urine, ascites cultures) before initiating corticosteroids regardless of fever presence 5

Step 3: Reassessment at Day 7

• Recalculate scores on day 7 to improve prognostic accuracy—this improves AUROC from 0.74-0.78 to 0.75-0.83 for 28-day mortality 6
• Calculate Lille score after 7 days of corticosteroid therapy 1
• Lille score ≥0.45: Indicates poor response with 75% 6-month mortality—stop corticosteroids and consider alternative therapy 1, 2
• Alternative assessment: A 25% fall in bilirubin after 1 week predicts improved survival (82% vs 44%, p=0.0005) with similar sensitivity to Lille score (90% vs 95%) 3

Step 4: Combined Score Approach for Optimal Prognostication

• Combining static (ABIC, GAHS, MELD) with dynamic (Lille) scores significantly improves outcome prediction compared to either alone (p<0.01 for all comparisons) 7
• MELD+Lille combination performs best with Akaike information criterion of 1305 versus 1313 for Maddrey+Lille and 1312 for ABIC+Lille 7
• Using combined GAHS/Lille approach reduces 90-day mortality to 20.6% compared to 26.8% with traditional DF ≥32 and Lille assessment 4

Comparative Performance and Evidence Quality

Superiority Over Traditional Scoring

• ABIC, GAHS, and MELD are superior to the discriminant function (DF) with AUROCs of 0.726,0.713, and 0.704 respectively versus 0.670 for DF at 90 days (p<0.05) 4
• The Korean Association for the Study of the Liver formally recommends mDF and MELD scores (A1 recommendation), while noting GAHS and ABIC "show potential" as prognostic models 1
• All models are good predictors of short-term mortality but poor predictors of long-term mortality 1

Validation Across Populations

• These scores have been validated in multiple international cohorts including France, United Kingdom, United States, Korea, and Belgium 7, 4, 6
• In population-based unselected cohorts, the models perform similarly with AUROC 0.74-0.78 for 28-day mortality, though slightly lower than in original derivation cohorts 6

Critical Clinical Pitfalls to Avoid

• Do not use GAHS or ABIC in isolation—combining with day 7 reassessment or Lille score substantially improves prognostic accuracy 7, 4
• Do not treat patients with GAHS <9 with corticosteroids—they have excellent outcomes (94% 28-day survival) without therapy and gain no benefit from treatment 3, 4
• Do not continue corticosteroids beyond 7 days if Lille score ≥0.45 or bilirubin fails to fall by 25%—these patients have 75% 6-month mortality and should have therapy stopped 1, 2, 3
• Do not overlook ABIC intermediate-risk patients (6.71-9.0)—this group specifically benefits from corticosteroid or pentoxifylline therapy 1
• Do not delay early prognostic stratification—calculating scores at presentation is crucial for timely intervention in severe disease with 30-50% untreated mortality 1, 5