Adjunctive fecal microbiota transplantation may reduce recurrent overt hepatic encephalopathy: a systematic review and Bayesian meta-analysis of three small randomized controlled trials

Recurrent overt hepatic encephalopathy (HE) in cirrhosis remains a major clinical challenge despite standard therapies including lactulose and rifaximin. Emerging evidence implicates gut dysbiosis in HE pathogenesis, positioning fecal microbiota transplantation (FMT) as a promising microbiome-restoring intervention. A systematic review and Bayesian meta-analysis were conducted to evaluate the efficacy and safety of adjunctive FMT in reducing HE recurrence in adults with cirrhosis. This Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-compliant review searched PubMed, Embase, CENTRAL, Scopus, and Google Scholar through September 2025 for randomized controlled trials (RCTs) comparing adjunctive FMT plus standard care versus standard care alone in adults with recurrent overt HE. Two independent reviewers performed screening, data extraction, risk-of-bias assessment (RoB 2), and Grading of Recommendations Assessment, Development and Evaluation (GRADE) certainty rating. Random-effects frequentist and hierarchical Bayesian meta-analyses with Markov Chain Monte Carlo sampling were performed. Three RCTs enrolling 70 participants (35 FMT, 35 standard care) met inclusion criteria; all three were conducted by the same principal investigator at a single center. For recurrent overt HE within 180 days, frequentist meta-analysis showed a reduction with adjunctive FMT (RR = 0.265, 95% CI 0.090–0.784, P = 0.016, I2 = 0%); however, leave-one-out sensitivity analysis showed that two of three iterations lost statistical significance, indicating a statistically fragile result. Bayesian analysis yielded a posterior median RR = 0.255 (95% credible interval [CrI] 0.057–1.039), with 97.3% posterior probability of benefit. Hospital readmission showed a nonsignificant trend toward reduction (RR = 0.422, 95% CI 0.167–1.066, P = 0.068). GRADE certainty was low for HE recurrence (downgraded for risk of bias from nonindependent, predominantly open-label trials, and for imprecision), low for readmission and SAEs, and very low for mortality. Wide prediction intervals indicate substantial uncertainty about future trial results. Preliminary evidence from three small RCTs suggests that FMT may reduce recurrent overt HE in cirrhosis, with no clear signal of increased harm; however, the Bayesian 95% credible interval crosses the null (0.057–1.039), leave-one-out analyses demonstrate fragility, and all three RCTs originate from a single investigator at one center. Independent multicenter replication is required before definitive conclusions can be drawn. Larger multicenter RCTs are needed to confirm these findings and establish optimal FMT protocols.