Acute hepatic inflammation from heavy alcohol use. AST typically 2–6× ULN with AST:ALT >2:1, jaundice, and recent heavy drinking. Maddrey Discriminant Function (DF) ≥32 defines severe disease -consider steroids. MELD score also predicts mortality. Nutrition may be the most important intervention.
🔍 Overview
Definition & Diagnosis
AST typically 2–6× ULN (rarely > 500 in pure alcoholic hepatitis)
AST:ALT ratio > 2:1 -classic pattern (alcohol damages mitochondrial AST, and ALT requires pyridoxal phosphate which is depleted in alcoholics)
Recent heavy alcohol use (typically > 40g/day women, > 60g/day men for > 6 months)
Jaundice (bilirubin often markedly elevated)
Exclude other causes: viral hepatitis, drug-induced, autoimmune
Maddrey Discriminant Function (DF)
DF = 4.6 × (patient PT − control PT) + total bilirubin
High-calorie, high-protein nutrition (35–40 kcal/kg/day) -nutrition may be the single most important intervention
Thiamine (Vitamin B1) 100 mg IV/PO daily
Folic acid 1 mg PO daily
Multivitamin daily
Alcohol cessation -absolute requirement
DF ≥ 32 -Severe
Prednisolone (Orapred) 40 mg PO daily × 28 days → taper
Calculate Lille score at day 7 -this is MANDATORY
Lille > 0.45 = non-responder → STOP steroids (no benefit, only side effects)
Lille < 0.45 = responder → complete 28-day course
PLUS all supportive care above (nutrition, thiamine, folate, MVI)
Contraindications to Steroids
Active GI bleed
Uncontrolled infection
HBV/HCV co-infection
Hepatorenal syndrome
Pentoxifylline for alcoholic hepatitis is DEAD.STOPAH, 2015 showed no benefit. Prednisolone showed modest short-term mortality benefit but no long-term benefit. Nutrition may be the most important intervention.
Updated Practice - Old teaching: pentoxifylline is an alternative to steroids. WRONG -STOPAH definitively showed pentoxifylline doesn't work. Steroids remain the only pharmacologic option for severe AH, but the benefit is modest and short-term.
🧪 Workup
Test
Purpose
AST, ALT
AST:ALT > 2:1 classic. AST rarely > 500.
Total/direct bilirubin
Often markedly elevated (used in DF and MELD)
INR/PT
Synthetic function, used in DF calculation
Albumin
Synthetic function and inflammation (NOT a nutritional marker, see below)
Why prednisolone and NOT prednisone for alcoholic hepatitis?
Prednisolone is the active form. Prednisone is a prodrug that requires hepatic conversion to prednisolone. In a patient with severe alcoholic hepatitis, the damaged liver cannot reliably convert prednisone → prednisolone. Using prednisolone bypasses this step and ensures the patient receives the active drug.
What is the Lille score cutoff and what does it mean?
Lille score > 0.45 at day 7 = non-responder. These patients derive no benefit from continued steroids and are only exposed to steroid side effects (infection, hyperglycemia). Stop steroids immediately. Lille < 0.45 = responder → complete the 28-day course. The Lille model uses day 0 and day 7 bilirubin, albumin, age, creatinine, and PT to predict response.
What did the STOPAH trial show?
STOPAH, 2015 was the largest RCT in alcoholic hepatitis (n=1103). Pentoxifylline showed NO benefit at any time point -it is definitively dead. Prednisolone showed a trend toward reduced 28-day mortality (OR 0.72, p=0.06) but no significant benefit at 90 days or 1 year. Bottom line: steroids provide modest short-term benefit; pentoxifylline provides none.
Why is AST:ALT ratio > 2:1 in alcoholic hepatitis?
Two reasons: (1) Alcohol damages mitochondria, which contain most of the AST (mitochondrial AST isoenzyme is released). ALT is primarily cytoplasmic. (2) Pyridoxal-5-phosphate (active B6) is depleted in alcoholics. ALT requires pyridoxal phosphate as a cofactor more than AST does, so ALT production is impaired. Result: AST rises more than ALT. An AST:ALT >2:1 is ~90% specific for alcoholic liver disease. If AST >500, consider other diagnoses (acetaminophen toxicity, ischemic hepatitis).
What is the role of early liver transplant in severe alcoholic hepatitis?
Traditionally, transplant programs required 6 months of sobriety before listing. However, Mathurin et al, NEJM 2011 showed that early transplant for steroid non-responders (Lille >0.45) improved 6-month survival from 23% to 77%. The ACCELERATE-AH, Hepatology 2021 confirmed these findings. Careful patient selection is key: first liver-decompensating event, strong social support, no other substance abuse, psychiatric evaluation. Relapse rates are comparable to patients transplanted after 6 months of sobriety.
What infections must you screen for before starting steroids?
Steroids in severe AH are immunosuppressive. Before starting, screen for: (1) Spontaneous bacterial peritonitis (SBP) - diagnostic paracentesis if ascites present (PMN >250 = SBP, contraindication to steroids until treated). (2) Blood cultures - active bacteremia. (3) CXR - pneumonia. (4) Urinalysis/culture - UTI. (5) HBV serologies - active HBV is a contraindication (steroids cause viral reactivation). Treat any active infection for 48-72h before starting prednisolone. During steroid therapy, maintain a low threshold to culture - steroids mask fever and inflammatory markers.
Why is nutrition so critical in alcoholic hepatitis?
Virtually all patients with severe AH are malnourished, even those who appear obese (sarcopenic obesity). Malnutrition independently predicts mortality - patients consuming <21.5 kcal/kg/day had 65% mortality vs 33% in those meeting caloric goals. Moreno et al, Gastroenterology 2019 Target: 35-40 kcal/kg/day with 1.2-1.5 g protein/kg/day. Use enteral nutrition (NG feeds) if unable to eat. Protein restriction for encephalopathy is OUTDATED - adequate protein is essential. Consult dietitian early.
How do you differentiate alcoholic hepatitis from other causes of jaundice in an alcoholic patient?
The differential for jaundice in an alcoholic includes: (1) Alcoholic hepatitis - AST:ALT >2:1, AST <500, recent heavy drinking, elevated bilirubin and INR. (2) Decompensated cirrhosis - may overlap; look for portal HTN signs (ascites, varices). (3) Acetaminophen toxicity - AST/ALT often >1000, check level. Alcoholics are at risk (depleted glutathione). (4) Biliary obstruction - dilated ducts on US, elevated ALP/GGT disproportionate to AST. (5) Hepatitis B/C - check serologies. (6) Ischemic hepatitis - AST/ALT >1000, recent hypotension. Liver biopsy is gold standard but rarely needed if clinical picture is classic.
What is the MELD score and how does it compare to Maddrey DF?
MELD = 3.78 x ln(bilirubin) + 11.2 x ln(INR) + 9.57 x ln(creatinine) + 6.43. Both predict short-term mortality: DF >=32 and MELD >21 identify severe AH. MELD has the advantage of including creatinine (captures hepatorenal syndrome) and is more objective (no "control PT" variability). Dunn et al, Hepatology 2005MELD >21 has ~75% sensitivity for 90-day mortality. In practice, calculate both. A high MELD with a Lille >0.45 at day 7 carries >75% 6-month mortality.
Clinical Examples
📋 Case 1 - Severe AH Responding to Steroids
Patient: 52M with 25-year history of heavy alcohol use (12 beers/day), presenting with 3 weeks of progressive jaundice, abdominal distension, and confusion. Last drink 5 days ago.
Labs: AST 340, ALT 120 (ratio 2.8:1), total bilirubin 22, INR 2.4, albumin 2.1, Cr 1.2, WBC 15.8, MCV 108. Hepatitis B/C negative.
Assessment: Severe alcoholic hepatitis (DF 81.8 >= 32). No contraindications to steroids (paracentesis: PMN 80, blood cultures negative, CXR clear).
Management: Prednisolone 40 mg PO daily. High-calorie nutrition (35 kcal/kg/day via oral + NG supplement). Thiamine 100 mg IV, folate 1 mg, MVI. Lactulose for encephalopathy. Day 7 Lille score: 0.18 (responder) - completed 28-day course. Bilirubin trended down to 8 by day 28. Discharged with alcohol cessation program referral.
📋 Case 2 - Severe AH, Steroid Non-Responder
Patient: 45F with daily vodka use x 10 years, presenting with jaundice, RUQ pain, fever 38.5C. AST 280, ALT 90, bilirubin 28, INR 2.8, Cr 1.8, albumin 1.9.
Management: Started prednisolone 40 mg daily. Day 7: bilirubin rose from 28 to 34 (worsening). Lille score = 0.72 (non-responder, >0.45). Prednisolone STOPPED immediately (no benefit, only infection risk). Continued aggressive nutrition. Hepatology consulted for early transplant evaluation given first liver-decompensating event, supportive family, and willingness to pursue sobriety. Listed and transplanted at day 21. Good outcome.
📋 Case 3 - Mild AH with Supportive Care
Patient: 38M presenting with jaundice, nausea, and RUQ tenderness. Drinks 6-8 beers daily x 5 years. AST 220, ALT 85, bilirubin 6.5, INR 1.3, Cr 0.9, albumin 3.2.
Assessment: Mild alcoholic hepatitis. Does NOT meet criteria for steroids (DF < 32).
Management: Supportive care only: high-calorie high-protein diet, thiamine 100 mg PO daily, folate 1 mg, MVI. Alcohol cessation counseling with motivational interviewing. Gabapentin for alcohol cravings. Hepatitis serologies negative. LFTs trended: bilirubin 6.5 -> 3.2 over 2 weeks. Discharged day 5 with outpatient GI follow-up and alcohol treatment program.
📣 Sample Presentation
One-Liner
"Mr. Davis is a 48-year-old man with 20-year history of heavy alcohol use (1 pint vodka daily) presenting with 2 weeks of jaundice, AST 280, ALT 95, bilirubin 18, INR 2.1, Maddrey DF 58 -severe alcoholic hepatitis."
Key Points to Cover on Rounds
Severe alcoholic hepatitis (DF 58). Started prednisolone 40 mg PO daily. Lille score to be calculated on day 7 -if > 0.45 will stop steroids. Hepatitis serologies negative. No active GI bleed, no uncontrolled infection. Nutrition: high-calorie high-protein diet (35 kcal/kg/day), thiamine 100 mg IV, folate 1 mg, MVI. Alcohol cessation counseling initiated. Monitoring: LFTs twice weekly, daily BMP for HRS surveillance, glucose QID on steroids.
Monitoring
Parameter
Frequency
Target / Action
Lille score
Day 7 (MUST calculate)
> 0.45 = stop steroids. < 0.45 = complete course.
LFTs
Twice weekly
Trending bilirubin, AST, ALT improvement
BMP
Daily
Creatinine (HRS surveillance), electrolytes
Infection screening
Ongoing
Steroids mask infection -low threshold to culture
Glucose
QID while on steroids
Steroids cause hyperglycemia -sliding scale insulin
Mental status
Each assessment
Hepatic encephalopathy monitoring
⚡ Summary
Recognize the Pattern
Jaundice and hepatic inflammation on a background of heavy drinking, usually within 8 weeks of the last drink.AST 2 to 6 times the upper limit with an AST:ALT ratio above 2:1, and AST rarely exceeds 500. A transaminase above 500 or an ALT-predominant pattern means look for another cause: viral hepatitis, drug injury, ischemia or acetaminophen.
Score the Severity
Maddrey discriminant function of 32 or more defines severe disease and is calculated from bilirubin and the prothrombin time prolongation. MELD also predicts mortality and is increasingly preferred. Severe untreated disease carries roughly 30 to 50% short-term mortality, which is why the score is calculated on admission rather than at discharge.
Steroids Are the Only Drug, and the Benefit Is Modest
Prednisolone 40 mg daily for 28 days in severe disease (DF 32 or above).STOPAH showed a short-term mortality benefit that did not persist at 90 days and 1 year.Use prednisolone, not prednisone, because conversion to the active form depends on a liver that is failing.
Pentoxifylline Is Dead
Old teaching offered it as an alternative to steroids; STOPAH showed definitively that it does not work. Do not use it, including in patients who cannot take steroids: there is no pharmacologic substitute, and the answer there is supportive care and transplant evaluation, not a drug that does nothing.
Check the Contraindications Before Starting Steroids
Active uncontrolled infection, active gastrointestinal bleeding, hepatorenal syndrome, and untreated sepsis.Screen for infection first, since it is present in a large minority at presentation and steroids given into it are lethal. Treat the infection, then reassess whether steroids are still indicated.
Use the Lille Score at Day 7
A Lille score above 0.45 after 7 days of prednisolone means non-response, so stop the steroid. Continuing beyond that point adds infection risk without benefit. Non-responders should be moved toward transplant evaluation, not toward a longer course.
Nutrition May Be the Most Important Intervention
These patients are profoundly malnourished, and low caloric intake predicts mortality independently of the discriminant function. Target 35 kcal/kg/day and 1.2 to 1.5 g/kg/day of protein, using enteral feeding if intake is inadequate. Do not restrict protein for encephalopathy, which is old teaching that worsens the catabolic state. Give thiamine before any glucose, plus folate and the other B vitamins, and watch for refeeding syndrome.
Do Not Read the Albumin as Nutrition
Albumin is a negative acute phase reactant and a marker of synthetic function and inflammation, not of nutritional status. A low albumin here reflects the liver and the inflammatory state, and it will not rise with feeding. Abstinence is the single strongest determinant of long-term survival, so alcohol use disorder treatment starts during the admission, and early transplant in carefully selected non-responders is now an accepted option rather than a fixed 6-month rule.