Dataset Information

Defining nosocomial transmission of Escherichia coli and antimicrobial resistance genes: a genomic surveillance study.

ABSTRACT:

Background

Escherichia coli is a leading cause of bloodstream infections. Developing interventions to reduce E coli infections requires an understanding of the frequency of nosocomial transmission, but the available evidence is scarce. We aimed to detect and characterise transmission of E coli and associated plasmids in a hospital setting.

Methods

In this prospective observational cohort study, patients were admitted to two adult haematology wards at the Cambridge University Hospitals NHS Foundation Trust in England. Patients aged 16 years and older who were treated for haematological malignancies were included. Stool samples were collected from study participants on admission, once per week, and at discharge. We sequenced multiple E coli isolates (both extended spectrum β-lactamase [ESBL]-producing and non-ESBL-producing) from each stool sample. A genetic threshold to infer E coli transmission was defined by maximum within-host single nucleotide polymorphism (SNP) diversity and the probability of drawing observed pairs of between-patient isolates at different SNP thresholds. Putative transmission clusters were identified when sequences were less than the genetic threshold. Epidemiological links for each transmission event were investigated. We sequenced all E coli positive blood samples from the two adult haematology wards.

Findings

We recruited 174 (51%) of 338 adult patients admitted to the wards between May 13 and Nov 13, 2015. We obtained and cultured 376 stool samples from 149 patients, of which 152 samples from 97 (65%) patients grew E coli. Whole-genome sequencing was done on 970 isolates. We identified extensive diversity in the bacterial population (90 sequence types) and mixed E coli sequence type carriage. 24 (26%) patients carried two sequence types, 12 (13%) carried three, and six (6%) patients carried four or more sequence types. Using a 17 SNP cutoff we identified ten clusters in 20 patients. The largest cluster contained seven patients, whereas four patients were included in multiple clusters. Strong epidemiological links were found between patients in seven clusters. 17 (11%) of 149 patients had stool samples positive for ESBL-producing E coli, the most common of which was associated with bla _CTX-M-15 (12 [71%] of 17). Long-read sequencing revealed that bla _CTX-M-15 was often integrated into the chromosome, with little evidence for plasmid transmission. Seven patients developed E coli bloodstream infection, four with identical strains to those in their stool; two of these had documented nosocomial acquisition.

Interpretation

We provide evidence of bacterial transmission and endogenous infection during routine care by integrating genomic and epidemiological data and by determining a genetic cutoff informed by within-host diversity in the studied population. Our findings challenge single colony-based investigations, and the widely accepted notion of plasmid spread.

Funding

UK Department of Health, Wellcome Trust, UK National Institute for Health Research.

SUBMITTER: Ludden C

PROVIDER: S-EPMC8410606 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Escherichia coli often carries linked antimicrobial resistance genes on transmissible genetic elements. Through coselection, antimicrobial use may select for unrelated but linked resistance or virulence genes. This study used unconditional statistical associations to investigate the relationships between antimicrobial resistance phenotypes and antimicrobial resistance genes in 151 E. coli isolates from healthy pigs. Phenotypic resistance to each drug was significantly associated with phenotypic resistance to at least one other drug, and every association found that the probability of observing the outcome resistance was increased by the presence of the predictor resistance. With one exception, each statistical association that was identified between a pair of resistance genes had a corresponding significant association identified between the phenotypes mediated by those genes. This suggests that associations between resistance phenotypes might predict coselection. If this hypothesis is confirmed, evaluation of the associations between resistance phenotypes could improve our knowledge of coselection dynamics and provide a cost-effective way to evaluate existing data until large-scale genotypic data collection becomes feasible. This could enable policy makers and users of antimicrobials to consider coselection in antimicrobial use decisions. This study also considered the unconditional relationships between resistance and virulence genes in E. coli from healthy pigs (aidA-1, eae, elt, estA, estB, fedA1, stx1, and stx2). Positive statistical associations would suggest that antimicrobial use may select for virulence in bacteria that may contaminate food or cause diarrhea in pigs. Fortunately, the odds of detecting a virulence gene were rarely increased by the presence of an antimicrobial resistance gene. This suggests that on-farm antimicrobial use did not select for the examined virulence factors in E. coli carried by this population of healthy pigs.

Project description:Surveillance of antimicrobial resistance (AMR) is an important component of public health. Antimicrobial drug use generates selective pressure that may lead to resistance against to the administered drug, and may also select for collateral resistances to other drugs. Analysis of AMR surveillance data has focused on resistance to individual drugs but joint distributions of resistance in bacterial populations are infrequently analyzed and reported. New methods are needed to characterize and communicate joint resistance distributions. Markov networks are a class of graphical models that define connections, or edges, between pairs of variables with non-zero partial correlations and are used here to describe AMR resistance relationships. The graphical least absolute shrinkage and selection operator is used to estimate sparse Markov networks from AMR surveillance data. The method is demonstrated using a subset of Escherichia coli isolates collected by the National Antimicrobial Resistance Monitoring System between 2004 and 2012 which included AMR results for 16 drugs from 14418 isolates. Of the 119 possible unique edges, 33 unique edges were identified at least once during the study period and graphical density ranged from 16.2% to 24.8%. Two frequent dense subgraphs were noted, one containing the five β-lactam drugs and the other containing both sulfonamides, three aminoglycosides, and tetracycline. Density did not appear to change over time (p = 0.71). Unweighted modularity did not appear to change over time (p = 0.18), but a significant decreasing trend was noted in the modularity of the weighted networks (p < 0.005) indicating relationships between drugs of different classes tended to increase in strength and frequency over time compared to relationships between drugs of the same class. The current method provides a novel method to study the joint resistance distribution, but additional work is required to unite the underlying biological and genetic characteristics of the isolates with the current results derived from phenotypic data.