Aminoglycosides, vancomycin, and metronidazole for people with cirrhosis and hepatic encephalopathy

RATIONALE: Hepatic encephalopathy is a common complication of cirrhosis. Its development is associated with increased morbidity and mortality. Its exact pathogenesis is unknown, but ammonia, produced by bacterial action in the intestine, plays a key role. Antibiotics modulate the gut flora and may reduce intestinal ammonia production. Aminoglycosides such as neomycin, paromomycin, and ribostamycin have been used to treat hepatic encephalopathy, as have other antibiotics such as vancomycin and metronidazole. OBJECTIVES: To assess the beneficial and harmful effects of aminoglycosides, vancomycin, and metronidazole versus placebo, no intervention, other antibiotics, or other active pharmacological interventions, for the prevention and treatment of hepatic encephalopathy in people with cirrhosis. SEARCH METHODS: We searched the Cochrane Hepato-Biliary Group Controlled Trials Register, CENTRAL, MEDLINE, Embase, and three other databases to 15 April 2025. We also searched online trials registries for ongoing and unpublished trials, undertook manual searches of meeting and conference proceedings, checked bibliographies of relevant articles, and corresponded with investigators and pharmaceutical companies. ELIGIBILITY CRITERIA: We included randomised clinical trials (RCTs) involving participants with cirrhosis and hepatic encephalopathy, or who were at risk of developing hepatic encephalopathy, comparing aminoglycosides, vancomycin, or metronidazole to (1) placebo or no intervention; or (2) other pharmacological agents, including non-absorbable disaccharides, other antibiotics, or other potentially beneficial agents (e.g. branched-chain amino acids, L-ornithine L-aspartate, nitazoxanide (a broad-spectrum antiparasitic/antiviral agent), and nicotinohydroxamic acid (a potent urease inhibitor). We included trials irrespective of publication status, outcomes reported, language, or blinding. We excluded trials involving people with hepatic encephalopathy associated with acute liver failure or with non-cirrhotic portal hypertension. OUTCOMES: The critical outcomes were all-cause mortality, hepatic encephalopathy, and serious adverse events. The important outcomes were non-serious adverse events and health-related quality of life (HRQoL). Our primary time point was the maximum length of follow-up. RISK OF BIAS: We used Cochrane's original risk of bias tool (RoB 1) to assess the risk of bias. SYNTHESIS METHODS: We used standard Cochrane methods. We undertook random-effects meta-analyses to calculate risk ratios (RRs) or standardised mean differences (SMDs), with 95% confidence intervals (CIs). We assessed heterogeneity with the I2 statistic, and the certainty of evidence with the GRADE framework. INCLUDED STUDIES: We included 24 RCTs, involving 1405 participants experiencing 1418 hepatic encephalopathy events. Twenty-three trials evaluated the treatment of hepatic encephalopathy and one, the secondary prevention of hepatic encephalopathy; we analysed these trials jointly. The trials assessed three aminoglycosides: neomycin (15 trials), paromomycin (three trials), and ribostamycin (one trial), as well as vancomycin (two trials), and metronidazole (three trials). Overall, 670 participants received these pharmacotherapies while 735 participants received a placebo or other potentially beneficial agents. We classified 22 of the 24 trials to be at an overall high risk of bias based on domain-level assessments. SYNTHESIS OF RESULTS: The certainty of evidence for all comparisons was low to very low, mainly due to risk of bias, imprecision, and heterogeneity. Twenty-three of the 24 trials, involving 1383 participants, reported all-cause mortality data. Aminoglycosides may increase mortality slightly compared to other potentially active agents (RR 1.64, 95% CI 1.03 to 2.62; I² = 0%; 3 studies, 166 participants). The evidence was very uncertain about whether aminoglycosides versus a placebo (RR 1.02, 95% CI 0.62 to 1.69; I² = 0%; 3 studies, 137 participants), non-absorbable disaccharides (RR 1.21, 95% CI 0.57 to 2.59; I² not applicable; 4 studies, 266 participants), or other antibiotics (RR 1.00, 95% CI 0.24 to 4.23; I² = 83%; 8 studies, 496 participants) result in a difference in mortality risk. The evidence was also very uncertain when comparing vancomycin to non-absorbable disaccharides (RR 0.94, 95% CI 0.26 to 3.40; I² not applicable; 2 studies, 72 participants), and metronidazole to other active agents (RR 0.97, 95% CI 0.14 to 6.66; I² = 0%; 3 studies, 242 participants). Nineteen trials involving 1281 participants reported data on hepatic encephalopathy. There may be little to no difference in the effects of aminoglycosides versus non-absorbable disaccharides (RR 0.84, 95% CI 0.67 to 1.05; I² = 0%; 3 studies, 251 participants), aminoglycosides versus other potentially active agents (RR 1.21, 95% CI 0.79 to 1.85; I² = 0%; 3 studies, 166 participants), and metronidazole versus other active agents (RR 1.50, 95% CI 0.89 to 2.54; I² = 48%; 2 studies, 208 participants). The evidence is very uncertain about the effect of aminoglycosides versus placebo, other antibiotics, and vancomycin versus non-absorbable disaccharides. Twenty trials, involving 1186 participants, reported a total of 328 serious adverse events. Aminoglycosides may slightly increase the risk of serious adverse events compared with other potentially active agents (RR 1.60, 95% CI 1.03 to 2.47; I² = 0%; 3 studies, 166 participants). The evidence is very uncertain when comparing aminoglycosides to placebo and other antibiotics, or when comparing vancomycin to non-absorbable disaccharides. Eighteen trials, involving 922 participants, reported a total of 96 non-serious adverse events. There may be a slight increase in the risk of adverse events when comparing aminoglycosides to placebo (RR 2.80, 95% CI 1.11 to 7.04; I² not applicable; 2 studies, 98 participants), and to other antibiotics (RR 3.24, 95% CI 1.08 to 9.70; I² = 0%; 8 studies, 251 participants). The evidence is very uncertain about the effects of aminoglycosides versus non-absorbable disaccharides or other active agents, and metronidazole versus other active agents. Only one trial assessed HRQoL, but reported the data in a form that precluded meta-analysis. Eight trials received support from pharmaceutical companies while six did not. Ten trials did not provide this information. AUTHORS' CONCLUSIONS: Due to low- or very low-certainty evidence, we do not know if aminoglycosides benefit hepatic encephalopathy compared to placebo or other potentially active agents. There may be a slight increase in the risks of mortality and serious adverse events with aminoglycosides compared to other agents, and of non-serious adverse events when compared to placebo and other antibiotics. We do not know if vancomycin or metronidazole improve clinically relevant outcomes. Only one trial assessed health-related quality of life. FUNDING: This Cochrane review received no specific funding. REGISTRATION: https://doi.org/10.1002/14651858.CD012734.