Effect of Prophylactic Antibiotics on Mortality in Severe Alcohol-Related Hepatitis: A Randomized Clinical Trial

Abstract

Importance: The benefits of prophylactic antibiotics for hospitalized patients with severe alcohol-related hepatitis are unclear.

Objective: To determine the efficacy of amoxicillin-clavulanate, compared with placebo, on mortality in patients hospitalized with severe alcohol-related hepatitis and treated with prednisolone.

Design, setting, and participants: Multicenter, randomized, double-blind clinical trial among patients with biopsy-proven severe alcohol-related hepatitis (Maddrey function score ≥32 and Model for End-stage Liver Disease [MELD] score ≥21) from June 13, 2015, to May 24, 2019, in 25 centers in France and Belgium. All patients were followed up for 180 days. Final follow-up occurred on November 19, 2019.

Intervention: Patients were randomly assigned (1:1 allocation) to receive prednisolone combined with amoxicillin-clavulanate (n = 145) or prednisolone combined with placebo (n = 147).

Main outcome and measures: The primary outcome was all-cause mortality at 60 days. Secondary outcomes were all-cause mortality at 90 and 180 days; incidence of infection, incidence of hepatorenal syndrome, and proportion of participants with a MELD score less than 17 at 60 days; and proportion of patients with a Lille score less than 0.45 at 7 days.

Results: Among 292 randomized patients (mean age, 52.8 [SD, 9.2] years; 80 [27.4%] women) 284 (97%) were analyzed. There was no significant difference in 60-day mortality between participants randomized to amoxicillin-clavulanate and those randomized to placebo (17.3% in the amoxicillin-clavulanate group and 21.3% in the placebo group [P = .33]; between-group difference, -4.7% [95% CI, -14.0% to 4.7%]; hazard ratio, 0.77 [95% CI, 0.45-1.31]). Infection rates at 60 days were significantly lower in the amoxicillin-clavulanate group (29.7% vs 41.5%; mean difference, -11.8% [95% CI, -23.0% to -0.7%]; subhazard ratio, 0.62; [95% CI, 0.41-0.91]; P = .02). There were no significant differences in any of the remaining 3 secondary outcomes. The most common serious adverse events were related to liver failure (25 in the amoxicillin-clavulanate group and 20 in the placebo group), infections (23 in the amoxicillin-clavulanate group and 46 in the placebo group), and gastrointestinal disorders (15 in the amoxicillin-clavulanate group and 21 in the placebo group).

Conclusion and relevance: In patients hospitalized with severe alcohol-related hepatitis, amoxicillin-clavulanate combined with prednisolone did not improve 2-month survival compared with prednisolone alone. These results do not support prophylactic antibiotics to improve survival in patients hospitalized with severe alcohol-related hepatitis.

Trial registration: ClinicalTrials.gov Identifier: NCT02281929.

Figure 1.. Participant Flow in a Trial of Prednisolone and Amoxicillin-Clavulanate for Alcohol-Related Hepatitis

MELD indicates Model for End-stage Liver Disease. See footnote d in Table 1 for description of MELD score calculation. ^aStratified by center. ^bThree patients who were randomized did not meet inclusion criteria: biopsy failure (n = 1), patient did not receive any treatment (n = 1), and patient received steroids within the last 6 months (n = 1). ^cTwo patients who were randomized did not meet the inclusion criteria: biopsy did not confirm alcohol-related hepatitis (n = 1) and patient had ongoing infection (n = 1). ^dVital status was unknown for 1 patient. ^eAll patients had known vital status. ^fThis patient was included at the beginning of the study, before the amendment to 90-day and 180-day survival was accepted.

Figure 2.. 180-Day Cumulative Incidence of All-Cause Mortality, 60-Day Cumulative Incidence of Infection, and 60-Day Cumulative Incidence of Hepatorenal Syndrome

Incidence of death was estimated using the Kaplan-Meier method by treating patients lost to follow-up without vital status information and patients who underwent liver transplant as censored at the last available follow-up. Median observation time was 180 (IQR, 95-180) days in the amoxicillin-clavulanate group and 180 (IQR, 80-180) days in the placebo group. Incidences of infection and hepatorenal syndrome were estimated using the Kalbfleisch and Prentice method by treating death as a competing event. There was no prespecified definition of infection in the study protocol. The diagnosis of infection was made by investigators based on standard medical management (summarized by Louvet et al).