Adding amoxicillin/clavulanate to prednisolone yields no survival benefit in severe alcoholic hepatitis

Roshini Claire Anthony
06 Aug 2021

The addition of amoxicillin/clavulanate to prednisolone does not extend survival among patients with severe alcoholic hepatitis, though it may reduce infection risk, according to a French study presented at ILC 2021.

Participants in the Antibiocor study were 284 patients aged 18–75 years with biopsy-proven alcoholic hepatitis, Maddrey’s discriminate function 32, and Model For End-Stage Liver Disease (MELD) score 21. They were randomized 1:1 to receive prednisolone (40 mg/day) in addition to oral amoxicillin (1 g)/clavulanate (125 mg) or placebo three times/day for 1 month. Prednisolone was not discontinued at 7 days if there was no response.

Baseline characteristics were similar in the amoxicillin/clavulanate and placebo arms including age (median 53.5 vs 52.5 years), male sex (71.8 percent vs 73.2 percent), bilirubin level (median 27.1 vs 31.1 mg/dL), international normalized ratio (median 2 in both groups), creatinine level (median 0.7 vs 0.8 mg/dL), Maddrey function (median 63.7 vs 69.9), and MELD score (median 24 vs 25).

Survival at 2 months in the intention-to-treat analysis did not significantly differ between patients in the amoxicillin/clavulanate and placebo groups (82.7 percent vs 78.1 percent; hazard ratio [HR], 0.769, 95 percent confidence interval [CI], 0.451–1.310; p=0.331), with 24 and 31 deaths occurring in the respective groups. [ILC 2021, abstract LBO-2631]

Survival outcomes between groups were also similar at 3 and 6 months (HR, 0.776; p=0.304 and HR, 0.874; p=0.519, respectively).

Infection rates at 2 months were lower in the amoxicillin/clavulanate than the placebo group (29.7 percent vs 41.5 percent; HR, 0.616, 95 percent CI, 0.417–0.909; p=0.015), with 42 and 59 events recorded in the amoxicillin/clavulanate and placebo groups, respectively.

Severe adverse events were reported in a comparable proportion of amoxicillin/clavulanate and placebo recipients (n=93 and 102, respectively), while severe infections occurred in 25 and 47 patients, respectively. Liver failure rates were also similar between groups (n=37 vs 36). One amoxicillin/clavulanate and three placebo recipients experienced severe Clostridium difficile infection, and one and four, respectively, severe diarrhoea. There were no incidents of drug-induced liver injury.

Hepatorenal syndrome incidence at 2 months was also comparable in the two groups (9.9 percent vs 9.2 percent [n=14 vs 13]; HR, 0.905; p=0.811), as was MELD score <17 (48.0 percent vs 54.95 percent; p=0.377). Response to treatment, as defined by Lille score <0.45, did not differ between the amoxicillin/clavulanate and placebo groups at day 7 (score: 0.43 vs 0.45; 56.74 percent vs 55.07 percent; p=0.78).

“In severe alcoholic hepatitis, prednisolone is the only treatment which improves 1-month mortality compared with placebo,” presented study author Professor Alexandre Louvet from the CHU de Lille, Lille, France.

However, prednisolone fails to demonstrate efficacy at 3 or 6 months, suggesting it is not an ideal drug, and paving the way for new strategies, he continued.

In addition, infection plays a key role in mortality among patients with severe alcoholic hepatitis, with about 25 percent of patients likely to develop an infection following initiating prednisolone treatment for this condition. [Gastroenterology 2009;137:541-548]

“[In this study,] a 30-day course of antibiotics [in addition to prednisolone] did not improve 2-month survival in patients with severe alcoholic hepatitis and MELD score 21,” said Louvet. However, antibiotic treatment reduced the likelihood of infections, he said.  


Editor's Recommendations