Higher antimicrobial gene burden in patients with cirrhosis vs. IBD, other diseases
Click Here to Manage Email Alerts
Patients with cirrhosis had a higher abundance of beta-lactamase, vancomycin and quinolone genes in their gut microbiota compared with patients with inflammatory bowel disease, chronic kidney disease and diabetes, according to a presentation from Digestive Diseases Week.
"What we found is alarming because it means that patients with cirrhosis may have a much higher burden of antibiotic resistance genes in their gut microbiome compared with other chronic diseases, which puts them at a greater risk for developing poor outcomes,” Jasmohan Bajaj, MD, MS, FACG, AGAF, told Healio Gastroenterology. “A combination of poor liver function, local and systemic immune response, repeated contacts with healthcare facilities and antibiotic use may be to blame."
Bajaj, associate professor of medicine, Virginia Commonwealth University and McGuire VA Medical Center, and colleagues enrolled 163 cirrhotic outpatients without IBD, primary sclerosing cholangitis or on dialysis; 37% of these patients had diabetes. None of the patients had other antibiotic exposure or hospitalization, according to Bajaj.
Researchers used metagenomics to assess stools collected and mapped to the Comprehensive Antibiotic Resistance Database for antimicrobial resistance gene families. They compared cirrhosis, IBD and chronic kidney disease (CKD) with available metagenomic data to determine the abundance, type and differences between groups. Investigators also examined diabetes plus cirrhosis vs. diabetes alone.
Bajaj and colleagues used permutational multivariate analysis of variance (PERMANOVA) /principal coordinate analysis and DESeq2 to compare antimicrobial resistance abundance between cirrhosis and IBD, CKD and diabetes. For all four studies, investigators compared individual genes using linear discriminant analysis effect size and false discovery rate-corrected Kruskal-Wallis tests. To determine separate diseases on antimicrobial resistance genes, they used machine learning with Random Forest analysis.
Results showed a significant difference among antimicrobial resistance gene abundance between cirrhosis and IBD, CKD and diabetes.
According to DESeq2, cirrhosis patients had a higher abundance of beta-lactamase, vancomycin and quinolone antimicrobial resistance genes compared with other diseases. Results from Kruskal-Wallis demonstrated higher vancomycin, quinolone, macrolide and beta-lactamase abundance in cirrhosis compared with other diseases.
Bajaj said he and his colleagues found an excellent separation in cirrhosis compared with other diseases on Random Forest Analysis with an area under the curve of 0.99, an accuracy of 0.98 and a true positivity rate in cirrhosis of 98%.
Perspective
Back to Top
Bajaj and colleagues examined the prevalence of antibiotic resistance genes present in stool in 163 patients with cirrhosis using metagenomics. They further mapped these results to a comprehensive antibiotic resistance database and compared their findings in the cirrhotic patients with other chronic disease populations including those with IBD, kidney disease and diabetes. None in the cirrhotic population had been exposed to antibiotics or hospitalized. The researchers’ analysis demonstrated that cirrhosis patients had higher rates of beta lactamase as well as vancomycin resistance and quinolone antimicrobial resistance genes in their microbiota compared with the other diseases studied.
These findings are significant as those with cirrhosis are at higher risk for infection and there are populations that require antibiotic prophylaxis. We also know that there are microbiome changes in the setting of cirrhosis where the presence of dysbiosis, which includes antibiotic resistance strains, can predict worse hepatic outcomes. It is likely that the impaired gut–liver axis contributed to the resistance patterns that is not present in other non-liver diseases. We will need to be careful in how we tailor and choose our antibiotics in these cirrhotic individuals and more data will be required in this population to help us truly determine when to introduce antibiotics, and what antibiotics to introduce. It is likely that choices of antibiotics are going to depend on local resistance patterns.
Collapse
Bajaj J. et al. Abstract 352. Presented at: Digestive Disease Week; May 21-24, 2021 (virtual meeting).
Click Here to Manage Email Alerts