Abstract
Acute pancreatitis (AP) is an inflammatory pancreatic disorder associated with significant morbidity. While gut microbiome dysbiosis has been implicated in AP pathogenesis, prior studies have predominantly focused on descriptive compositional changes rather than linking specific microbial signatures to clinical inflammatory markers. This study aims to bridge this gap by identifying gut microbiome biomarkers correlated with systemic inflammation in AP. Fecal samples were collected from 34 patients with AP (diagnosed within 48 h of onset) and 20 age-/gender-matched healthy controls (HC). High-throughput 16S rRNA sequencing analyzed taxonomic profiles, while a random forest algorithm was employed to construct a diagnostic model based on differentially abundant species. Spearman correlation analysis assessed associations between key microbial taxa and inflammatory indicators (e.g., white blood cell count, procalcitonin). Patients with AP exhibited reduced gut microbiome diversity (Shannon index: 4.28 ± 1.02 vs. 4.93 ± 0.99, p = 0.027) and distinct taxonomic shifts, including enrichment of Fusobacterium periodonticum, Fusobacterium mortiferum, and Veillonella parvula (all p < 0.05), alongside depletion of Bacteroides plebeius and Faecalibacterium prausnitzii. A three-species diagnostic model (Veillonella parvula, Bacteroides plebeius, and Fusobacterium periodonticum) achieved an area under the ROC curve (AUC) of 0.94 (95% CI 0.88-1.00), with 88.2% sensitivity and 95.0% specificity. Notably, Veillonella parvula and Fusobacterium periodonticum abundances correlated positively with white blood cell count (r = 0.40, p = 0.018) and procalcitonin (r = 0.41, p = 0.015). Functional prediction revealed elevated microbial pathways linked to cell membrane transport and infectious disease risk in AP. This study confirms compositional and functional gut microbiome alterations in AP, with specific taxa strongly associated with inflammatory biomarkers. The validated three-species model demonstrates potential as a non-invasive diagnostic tool, advancing translational applications of microbiome profiling in AP management.