基因组
微生物群
肝硬化
生物
屎肠球菌
疾病
肠道微生物群
抗生素耐药性
肠道菌群
医学
人类微生物组计划
人体微生物群
抗生素
免疫学
病因学
微生物学
肠球菌
代谢组
生物信息学
小桶
内科学
计算生物学
作者
Darja Nikitina,Ki Tae Suk,Aldo Torre,Ramazan Idılman,Mario R Alvares-da-Silva,Mette Munk Lauridsen,Andrew Fagan,Edita Kreivėnaitė,Satya Priya Sharma,Elise Jonasson Nielsen,Dilara Turan,Vinay Jahagirdar,Matheus Truccolo Michalczuk,Alexander Khoruts,Sai Hruthik Bojja,Hye Won Lee,Jurgita Skiecevičienė,Masoumeh Sikaroodi,Juozas Kupcinkas,Patrick Martin Gillevet
出处
期刊:Gut
[BMJ]
日期:2026-09-21
卷期号:: gutjnl-2026
标识
DOI:10.1136/gutjnl-2026-339378
摘要
Background Hospitalisations represent major clinical events in cirrhosis, yet prediction based on clinical variables alone remains limited. Objective Given the role of the gut microbiome in disease progression, we evaluated whether gut metagenomic profiles are associated with 90-day hospitalisation and provide additional prognostic information beyond clinical features in a multinational outpatient cirrhosis cohort. Design We enrolled 679 outpatients with cirrhosis from seven countries and performed stool metagenomic profiling, including taxonomic, functional pathway and antibiotic resistance gene (ARG) analyses, with 90-day follow-up. Machine learning models were developed using clinical and microbiome features for predicting non-elective hospitalisations. Results 25% of patients required hospitalisation within 90 days. Hospitalised patients had more advanced cirrhosis, lower microbial diversity which was consistent across countries despite marked variation in microbial composition. Diet had a modest influence on microbiome structure. After adjustment for country, disease severity, cirrhosis aetiology and treatment, 19 bacterial species remained independently associated with hospitalisation, including enrichment of Enterococcus faecium and Veillonella rogosae and depletion of multiple commensal taxa. Functional profiling demonstrated coordinated taxonomic-functional alterations focusing on complex carbohydrate degradation pathways, glycan biosynthesis, lipid and nucleotide salvage pathways and association with antimicrobial resistance mechanisms. The combined clinical–microbiome model significantly outperformed both the clinical-only (area under the curve (AUC) 0.79) and microbiome-only (AUC 0.74) models, achieving an AUC of 0.84. Conclusion Gut microbiome composition and function are associated with short-term hospitalisation risk and provide additional prognostic information beyond clinical variables in a multicountry cohort. Hospitalisation, regardless of country, is characterised by loss of short-chain fatty acid-producing taxa, functional shifts, ARG and pathway changes.
科研通智能强力驱动
Strongly Powered by AbleSci AI