STOPAH

Published

Thursz MR, et al. "Prednisolone or pentoxifylline for alcoholic hepatitis". The New England Journal of Medicine. 2015. 372(17):1619-1628.
PubMed • Full text • PDF

Clinical Question

In patients with severe alcoholic hepatitis, does treatment with prednisolone or pentoxifylline reduce mortality at 28 days?

Bottom Line

Among patients with severe alcoholic hepatitis, neither prednisolone nor pentoxifylline reduces all-cause mortality at 28 days. Prednisolone was associated with a non-significant mortality benefit at 28 days but no benefit at 90 days or 1 year.

Major Points

The limited research into effective therapies for alcoholic hepatitis has primarily focused on prednisolone and pentoxifylline. For prednisolone, a 2008 Cochran review showed a trend towards benefit with glucocorticoid treatment. This was supported in a 2011 meta-analysis of the five largest studies, which showed a mortality benefit with prednisolone treatment. The evidence for pentoxifylline is less well established, with meta-analyses failing to show benefit. The evidence supporting the use of pentoxifylline comes from a single trial showing improved mortality with administration (see Pentoxifylline in Severe Alcoholic Hepatitis, 2000).

Published in 2015, the Steroids or Pentoxifylline for Alcoholic Hepatitis (STOPAH) trial randomized 1,092 patients with alcoholic hepatitis at 65 UK centers to a combination of prednisolone, pentoxifylline, and placebo in a double-blinded, 2x2 factorial design. Neither therapy showed significant reduction in mortality at 28 days and there were no significant differences between the mortality or liver transplantation rates at 90 days or at 1 year. (However, the prednisolone was associated with a mortality benefit trend of OR 0.72 with P=0.06.) A logistic regression suggests that prednisolone may provide a mortality benefit at 28 days, though not at 90 or 365 days. Further, prednisolone was associated with an increase in incident infections.

Guidelines

AASLD Alcohol-Associated Liver Disease (2019, adapted)

• If severe alcoholic hepatitis (defined by Maddrey's discriminant function score ≥32):
  • Administer prednisolone 40 mg/day PO if no corticosteroid contraindications.
  • Adding IV NAC to prednisolone might improve 30 day survival
  • Use the Lille score to reassess prognosis, guide treatment course after 7 days of corticosteroids, and identify non-responders.
• Assess malnutrition if alcoholic hepatitis.
• Encourage alcohol abstinence.
• No longer recommend the use of pentoxifylline in management of alcoholic hepatitis

Design

• Prospective multicenter randomized, double-blind 2x2 factorial trial
• N=1,092 (5,234 screened)
  • Placebo-placebo (n=272)
  • Prednisolone-placebo (n=274)
  • Pentoxifylline-placebo (n=273)
  • Prednisolone-pentoxifylline (n=273)
• Setting: 65 hospitals across the United Kingdom
• Enrollment: 2011-2014
• Follow-up: 12 months (primary outcome at 28 days)
• Analysis: Intention-to-treat
• Primary outcome: All-cause mortality at 28 days

Population

Inclusion Criteria

• Age ≥18
• Clinical diagnosis of alcoholic hepatitis
• Average alcohol consumption of:
  • >80 g/day for males
  • >60 g/day for females
• Serum bilirubin >80 μmol/L (4.7 mg/dL)
• Maddrey's discriminant function ≥32

Exclusion Criteria

• Jaundice for >3 months
• Cessation of alcohol consumption >2 months prior to randomization
• Other causes of liver disease
• Serum AST >500 IU/L or ALT >300 IU/L
• Development of renal failure, active GI bleeding, untreated sepsis or requirement of inotrope support, unless it becomes stable in the first week of admission

Baseline Characteristics

From the placebo-placebo group.

• Demographics: Age 49 years, male 60%
• Duration from admission to treatment: 6.1 days
• Alcohol consumption (g/day): Women 154, men 195
• Encephalopathy:
  • None: 70%
  • Grade:
    • 1: 17%
    • 2: 7%
    • 3: 2%
    • 4: 0%
• Conditions on admission: Sepsis 11%, renal failure <0.5%, GI bleeding 6%
• Labs: Bilirubin 18 mg/dL, albumin 26 g/L, AST 144 U/L, alk phos 185 U/L, creatinine 0.83 mg/dL, WBC count 10.1K cells/mm³, PMNs 7.6K cells/mm³, PT 21 seconds
• Scores: Discriminant function 62, MELD 21, GAHS 8

Interventions

• Participants were randomized to a group:
  • Placebo-placebo
  • Prednisolone-placebo
  • Pentoxifylline-placebo
  • Prednisolone-pentoxifylline
• Medications doses:
  • Prednisone 40 mg daily
  • Pentoxifylline 400 mg TID
• Study medications were continued for 28 days

Outcomes

Presented as prednisolone vs. no prednisolone vs. pentoxifylline vs. no pentoxifylline

Primary Outcome

All-cause mortality at 28 days

14% vs. 18% vs. 16% vs. 16%

Prednisolone: OR 0.72; 95% CI 0.52-1.01; P=0.06

Pentoxifylline: OR 1.07; 95% CI 0.77-1.49; P=0.69

Secondary Outcomes

All-cause mortality or liver transplant at 90 days

30% vs. 29% vs. 29% vs. 30%

Prednisolone: OR 1.02; 95% CI 0.77-1.35; P=0.87

Pentoxifylline: OR 0.97; 95% CI 0.73-1.28; P=0.81

All-cause mortality or liver transplant at 1 year

57% vs. 56% vs. 56% vs. 57%

Prednisolone: OR 1.01; 95% CI 0.76-1.35; P=0.94

Pentoxifylline: OR 0.99; 95% CI 0.74-1.33; P=0.97

Additional Analyses

Association of all-cause mortality and receipt of prednisolone vs. no prednisolone

Logistic regression with selection of included dependent variables using backward elimination (P<0.05). As selection of dependent variables was not decided a priori, the WJC editors consider this to be an exploratory analysis. No univariate analysis was presented.

At 28 days: OR 0.61; 95% CI 0.41-0.91; P=0.02

At 90 days: OR 1.00; 95% CI 0.73-1.36; P=0.98

At 1 year: OR 1.01; 95% CI 0.74-1.39; P=0.94

Association of all-cause mortality and receipt of pentoxifylline vs. no pentoxifylline

Same methods as above.

At 28 days: OR 1.10; 95% CI 0.74-1.64; P=0.62

At 90 days and at 1 year: Not presented

Adverse Events

Infection

Prednisolone vs. no-prednisolone: 13% vs. 7% (P=0.002)

Lung infection: 7% vs. 3% (P=0.007)

Criticisms

• The trial was stopped prematurely, allowing for collection of all 28 days follow-up, so 33 individuals were not included in 90 day or 1 year follow-up and another 159 individuals could not be included in the 1 year follow-up
  • Goal enrollment was 1,026, which the investigators met
• Lower mortality than expected
• Use of discriminant function without liver biopsy probably led to many misclassifications
• No taper in the prednisolone may have resulted in harm to the patients upon medication withdrawal

Funding

• NIHR grant from the Health Technology Assessment program