Global Carriage of Antimicrobial Resistance Genes and Virulence Factors Among Physicians in Cirrhosis Care

Technical Summary: Global Carriage of Antimicrobial Resistance Genes and Virulence Factors Among Physicians in Cirrhosis Care

Problem Statement
Patients with liver cirrhosis are highly susceptible to drug-resistant infections due to gut microbiome dysbiosis, immune impairment, and frequent hospital exposure. Infections caused by opportunistic pathogens such as Enterobacteriaceae, Enterococcus, and Klebsiella pneumoniae are associated with high mortality. While healthcare settings are recognized as reservoirs for antimicrobial resistance genes (ARGs), the extent to which physicians themselves serve as carriers of clinically relevant ARGs and virulence factors (VFs) remains under-characterized, particularly across diverse global healthcare settings. Understanding whether physicians act as primary reservoirs for these resistance determinants is critical for infection control strategies in cirrhosis care.

Methodology
The study employed shotgun metagenomic sequencing to analyze fecal samples from a global cohort of 55 physicians caring for cirrhosis patients across nine countries. The cohort was stratified by country income level: 42 physicians from high-income countries (HICs) and 13 from low- and middle-income countries (LMICs). Physicians were further categorized based on their exposure levels to inpatient/outpatient cirrhosis cases and patients with known resistant infections.

Additionally, a paired analysis was conducted within the United States (Virginia Commonwealth University and Richmond VA Medical Center), comparing the fecal metagenomes of 8 physicians against 172 cirrhosis patients (70 compensated, 102 decompensated) from the same facilities.

Data Processing and Analysis:

• Sequencing: Samples underwent shotgun metagenomic sequencing (median depth: 31 million reads for physicians, 19 million for patients).
• Bioinformatics: Reads were processed using KneadData for quality control and host removal. Taxonomic profiling was performed using MetaPhlAn 4.
• ARG and VF Identification: ARGs and VFs were identified via read-based alignment against the ResFinder and VFDB databases using KMA, with hits filtered for ≥90% adjusted coverage.
• Assembly-Based Analysis: A parallel assembly approach (MEGAHIT, METABAT) was used to reconstruct metagenome-assembled genomes (MAGs) to link ARGs and VFs to specific microbial hosts. Plasmid replicons were screened using PlasmidFinder.
• Statistical Analysis: Differences in abundance and prevalence were assessed using Wilcoxon rank-sum tests, Fisher's exact tests, and generalized linear models, with adjustments for multiple comparisons (Benjamin-Hochberg).

Key Results

1. Global ARG Carriage in Physicians:

   • Physicians from LMICs exhibited significantly higher ARG richness (diversity) compared to those from HICs, a finding correlated with higher species richness in the gut microbiome.
   • However, the total normalized abundance of ARGs did not differ significantly between income groups.
   • Clinically relevant ARGs (e.g., vancomycin resistance, ESBLs like blaCTX, quinolone resistance) were rare across all physician samples. When detected, they were often associated with specific individuals reporting recent infections rather than general practice exposure.
   • Most detected ARGs belonged to commensal taxa (e.g., Bacteroidaceae, Lachnospiraceae) rather than high-risk pathogens.

2. Impact of Patient Exposure:

   • While physicians with high exposure to patients with resistant infections showed a trend toward higher ARG counts, the differences were not statistically significant for total abundance.
   • Specific genes like erm(X) were more prevalent in high-exposure groups, but no strong correlation was found between exposure metrics and the carriage of high-risk clinical resistance determinants.

3. Virulence Factors (VFs):

   • The prevalence of detectable E. coli was significantly higher in physicians from LMICs (76.9%) compared to HICs (40.5%).
   • K. pneumoniae and Enterococcus were rarely detected in physicians.
   • While 16 physicians carried detectable Enterobacteriaceae-associated VFs (including adhesion factors and siderophores), the burden was generally low and did not strictly correlate with patient exposure or country of practice.

4. Physician vs. Cirrhosis Patient Comparison (US Cohort):

   • Cirrhosis patients carried a substantially greater burden of clinically relevant ARGs and VFs compared to their paired physicians.
   • Key high-risk ARGs (vancomycin resistance, quinolone resistance, and specific β-lactamases like blaCTX, blaSHV, blaTEM) were unique to the cirrhosis patient cohort and absent in physician samples.
   • While physicians showed higher species richness, patients exhibited a distinct enrichment of specific resistance genes, such as tet(46) on Streptococcus MAGs, likely driven by oral microbiome translocation.
   • Plasmid-linked ARGs were detected in both groups but were more prevalent and diverse in the patient cohort.

Significance and Claims
The study concludes that while physicians in cirrhosis care harbor a background resistome that varies by geography and microbial diversity, they are unlikely to serve as primary reservoirs for clinically relevant, high-risk antimicrobial resistance genes and virulence factors. The "high-risk" resistance profile observed in cirrhosis patients—characterized by specific ESBLs, vancomycin resistance, and virulence factors—is distinct from that of the treating physicians.

The authors emphasize that the observed differences in ARG carriage are largely driven by underlying microbial diversity (higher in LMICs) rather than direct transmission from patients to physicians. Consequently, the study suggests that infection control efforts should focus on the patient microbiome and environmental reservoirs rather than assuming physicians are the primary vectors for drug-resistant pathogens in this setting. However, the authors note that individual physicians could still play a role in bidirectional transmission, warranting continued global surveillance. The study acknowledges limitations, including the small sample size of the physician cohort and the reliance on self-reported exposure data, but asserts that these findings provide a necessary baseline for understanding ARG dissemination in cirrhosis care.