Prednisolone in Severe Alcoholic Hepatitis

Published

Ramond MJ, et al. "A randomized trial of prednisolone in patients with severe alcoholic hepatitis". The New England Journal of Medicine. 1992. 362(8):507-512.
PubMed • Full text • PDF

Clinical Question

In patients with severe alcoholic hepatitis, does prednisolone improve survival?

Bottom Line

In patients with severe alcoholic hepatitis, prednisolone improves 2-month survival.

Major Points

Alcoholic hepatitis carries up to a 65% mortality rate, and because therapy generally has consisted mainly of supportive measures, many investigators have sought out disease-modifying interventions. In a sea of studies yielding conflicting results, the present 1992 study was one of the first to suggest a survival benefit of glucocorticoids among patients with severe alcoholic hepatitis.

This trial randomized 61 patients with severe, biopsy-proven alcoholic hepatitis to prednisolone or placebo. Severe disease was determined by the presence of hepatic encephalopathy or Maddrey's discriminant function (DF) >32. At 2 months, survival was significantly better in patients receiving prednisolone (88% vs. 45%) with a NNT of 3. This survival benefit was also seen at 6 months. In contrast to other trials, the prednisolone group experienced fewer GI bleeds and infections, which one would have expected to be more common in the glucocorticoid group.

Since the publication of this trial, two positive meta-analyses have again underscored that steroids are most beneficial among those with severe alcoholic hepatitis, manifested by encephalopathy or DF >32. A 2008 Cochrane Review suggests that there is insufficient data to support or advise against the routine use of steroids in alcoholic hepatitis. The 2009 AASLD/ACG guidelines recommend consideration of prednisolone for patients with severe alcoholic hepatitis (based upon the presence of hepatic encephalopathy or DF ≥32), and supportive treatment alone for those with mild to moderate disease.

The subsequent Pentoxifylline in Severe Alcoholic Hepatitis (2000) suggested that pentoxifylline may also improve survival in severe alcoholic hepatitis. The 2015 STOPAH trial randomized patients to pentoxifylline, prednisolone, or placebo. There was no mortality benefit with pentoxifylline, but prednisolone was associated with a non-significant trend towards mortality reduction at 28 days (OR 0.72; 95% CI 0.52-1.01; P=0.06).

Guidelines

AASLD/ACG Alcoholic Liver Disease (2010, adapted)

• For patients with alcoholic hepatitis:
  • Alcohol abstinence counseling (class I, level B)
  • Assessment for nutritional, vitamin, and mineral deficiencies with aggressive repletion in those with severe disease (class I, level B)
  • If mild-moderate (Maddrey score <32), no hepatic encephalopathy, and improvement in bilirubin or decline in Maddrey score, unlikely to benefit from medical interventions except abstinence and nutritional support (class III, level A)
  • If severe (Maddrey score ≥32) regardless of hepatic encephalopathy, if no conraindications to corticosteroids, consider four weeks of prednisone at 40 mg/day for 28 days then discontinuation or tapering dose over two weeks (class I, level A)
  • If severe (Maddrey score ≥32), consider pentoxifylline 400 mg PO TID for 28 days, especially if corticosteroid contraindications (class I, level B)

Design

• Double-blind, parallel group, randomized, placebo-controlled trial
• N=61
  • Prednisolone (n=32)
  • Placebo (n=29)
• Setting: two centers in France
• Enrollment: 1987-1990
• Follow-up: 6 months
• Analysis: not specified
• Primary end point: 60-day survival

Population

Inclusion Criteria

• Chronic alcoholism and clinical features of alcoholic hepatitis
• Biopsy-proven alcoholic hepatitis, characterized by hyaline necrosis and infiltration of polymorphonuclear leukocytes
• Hepatic encephalopathy and/or Maddrey's discriminant function >32

Exclusion Criteria

• Gastrointestinal bleeding
• Bacterial infection
• Liver biopsy negative for alcoholic hepatitis
• Gastric or duodenal ulcer
• Malignancy
• Presence of HBsAg
• Presence of HIV antibodies
• Anticoagulation therapy

Baseline Characteristics

• Age: 48 years
• Female: 69%
• Ascites: 80%
• Encephalopathy: 31%
• Esophageal varices: 51%
• Hepatic venous pressure gradient: 19.5 mmHg
• Bilirubin: 14.4 mg/dL
• PT: 38% normal
• AST: 3.5 x ULN
• Albumin: 402 umol/L
• Hgb: 6.4 mmol/L
• Discriminant function: 55

Interventions

• HBV and HIV serologies, gastroscopy, liver biopsy
• Discriminant function: 4.6 (PT - control time in seconds) + serum bilirubin (mmol/L)/17
• Randomized to either prednisolone 40 mg or placebo by mouth every morning for 28 days.
• Routine treatment regimen included:
  • 3,000 cal diet with 1g protein/kg TBW
  • Lactulose for hepatic encephalopathy
  • Sodium restriction or spironolactone for ascites
  • Fluid restriction for hyponatremia
  • B-complex vitamins, folate, and antacids supplementation

Outcomes

Comparisons are prednisolone vs. placebo.

Primary Outcome

60-day survival

88% vs. 45% (P=0.001; NNT=3)

Secondary Outcomes

30-day survival

88% vs. 62%

6-month survival

84% vs. 45% (P=0.002; NNT=3)

Subgroup Analysis

• Survival advantage for prednisolone persisted after stratification according to center and presence of encephalopathy, and after adjustment for prognostic factors.

Adverse Events

• Worsening of psychiatric disorders and glucose intolerance was noted in one member of the prednisolone group.
• Infection and GI bleeding were more common in the placebo group.

Criticisms

• Delay from admission to randomization in order to get liver biopsy
• Patients in the placebo group may have more severe disease with higher baseline creatinine, bilirubin and DF.
• Nearly 70% women renders the results difficult to generalize. In US, 60-70% patients with alcoholic hepatitis are male. In addition, some studies have demonstrated that females may have more favorable responses to glucocorticoids.

Funding

Not identified