Corticosteroids in severe alcohol-related hepatitis: a multicenter randomized clinical trial.

[en] BACKGROUND & AIMS: Severe alcohol-related hepatitis (AH) has high short-term mortality, and corticosteroid therapy is recommended in the absence of contraindications. However, its benefit in patients with early spontaneous bilirubin improvement remains unknown. We determined whether corticosteroid therapy improves survival compared with placebo in this specific population.
METHODS: In this multicenter, randomized, placebo-controlled trial conducted in 10 Belgian hospitals (February 2018-May 2024), patients aged ≥18 years with biopsy-proven severe AH and spontaneous bilirubin decline >10% between admission and day 5-10 were randomized 1:1 to methylprednisolone 32 mg daily or placebo for 28 days. The primary outcome was 90-day survival; other outcomes included 30-day survival and cumulative incidence of infection at 90 days.
RESULTS: Sixty-nine of the 140 planned patients were analyzed (38 corticosteroid, 31 placebo), representing 49% of the planned sample size. Baseline characteristics were well balanced. At 90 days, survival was 79% (95% CI, 67%-93%) versus 83% (95% CI, 70%-98%) (hazard ratio, 1.19; 95% CI, 0.39-3.63; P=0.76). Eight deaths occurred in the corticosteroid arm versus 5 in the placebo arm. At 30 days, survival was 95% versus 94% (P=0.83). Cumulative incidence of infection at 90 days was 47% versus 34% (subdistribution hazard ratio, 1.55; 95% CI, 0.72-3.34; P=0.26). Lille score was an independent predictor of 90-day mortality.
CONCLUSIONS: In patients with severe AH and early spontaneous bilirubin improvement, corticosteroid therapy was not associated with improved survival, but the underpowered trial cannot exclude a benefit. A short observation period after admission may help identify patients who can be managed without corticosteroids.
IMPACT AND IMPLICATIONS: Corticosteroid therapy is recommended for patients with severe alcohol-related hepatitis (AH), but whether this benefit extends to the subgroup who show early spontaneous improvement in bilirubin after admission was previously unknown. In this multicenter, randomized, placebo-controlled trial, corticosteroid therapy was not associated with improved 90-day or 30-day survival in patients with severe, biopsy-proven AH and a spontaneous bilirubin decline greater than 10% within 5 to 10 days of admission, and was associated with a numerically higher, though not statistically significant, risk of infection. These findings are most relevant to hepatologists and other clinicians managing patients with severe AH, and to researchers designing future AH trials, as they identify a subgroup in whom the benefit of corticosteroids is uncertain. They support a risk-stratified approach in which a brief observation period after admission-incorporating bilirubin trend, infection screening, and liver biopsy-may help identify patients with a relatively favorable prognosis who can be spared corticosteroid therapy and its associated risks. However, because the trial was stopped early after enrolling only 69 of the 140 planned patients, it was underpowered, and a modest survival benefit of corticosteroids cannot be excluded. These results should therefore not be interpreted as definitive evidence against corticosteroid use in this population and require confirmation in larger, adequately powered studies before informing clinical guidelines or being generalized beyond this specific patient subgroup.
CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov (NCT03160651), EudraCT number: 2016-005136-16.