S271 The Efficacy and Safety of Microbiome Modulating Agents Alongside Standard Management for Acute Pancreatitis: An Updated Systematic Review and Network Meta-Analysis

Introduction: Acute pancreatitis (AP) admissions continue to rise with gall-stone disease and alcohol use. Contemporary protocols have cut mortality and length-of-stay (LOS), yet inflation-adjusted hospital costs have doubled. Bacterial translocation across a disrupted gut barrier is the main driver of infected necrosis and sepsis, prompting interest in microbiome-modulating agents (MMAs: probiotics or synbiotics). These products are costly and not standard of care. We therefore synthesized all modern evidence to clarify whether MMAs improve clinically relevant AP outcomes. Methods: Embase, MEDLINE, CENTRAL and Scopus were searched to 8 May 2025. Two reviewers screened and extracted data; high-risk studies (RoB-2 or Newcastle-Ottawa) were removed in sensitivity analyses. A Bayesian random-effects network meta-analysis (gemtc, R2jags 4.3.2) pooled direct and indirect evidence; node-splitting tested inconsistency and meta-regression explored severity effects. The primary endpoint was a composite of infected pancreatic necrosis and extra-pancreatic infections; secondary endpoints were all-cause mortality and LOS. Odds ratios (OR) were calculated for dichotomous and mean differences for continuous data. Results: Nineteen studies (15 randomized controlled trials, 4 cohorts; n = 1 430, 83% severe AP) met criteria and were included in our review. Three probiotic/synbiotic regimens formed a connected evidence network.4-strain Chinese mix + early enteral nutrition vs control: infection OR 0.28 (0.11–0.69), I² 12%; LOS –5.6 days (–9.2 to –2.1); mortality OR 0.28 (0.06–1.29). Synbiotic 2000: infection OR 0.41 (0.18–0.93), I² 5%; LOS –4.9 days (–8.1 to –1.7); mortality OR 0.50 (0.11–2.27). PROPATRIA 6-strain mix: infection OR 0.96 (0.67–1.37); LOS mean differences +0.3 days (–2.4 to +3.1); mortality excess OR 2.53 (1.22–5.25). An indirect comparison of Synbiotic 2000 vs 4-strain mix showed no significant difference (OR 1.46 [0.55–3.86]). Global heterogeneity was low–moderate (I² ≤ 38%); no inconsistency or severity interaction was detected. Conclusion: Modern synbiotic formulations (Synbiotic 2000 and the contemporary 4-strain Chinese mix) reduce infectious complications and shorten hospital stay in AP without increasing deaths, whereas the older PROPATRIA preparation confers no benefit and may be harmful. Benefits appear strain-specific; MMAs should not be used interchangeably. Large, multinational randomized controlled trials evaluating a single, standardized agent are warranted before routine adoption.