Genotypic resistance determined by whole genome sequencing versus phenotypic resistance in 234 Escherichia coli isolates.

Whole genome sequencing (WGS) enables detailed characterization of bacteria at single nucleotide resolution. It provides data about acquired resistance genes and mutations leading to resistance. Although WGS is becoming an essential tool to predict resistance patterns accurately, comparing genotype to phenotype with WGS is still in its infancy. Additional data and validation are needed. In this retrospective study, we analysed 234 E. coli isolates from positive blood cultures using WGS as well as microdilution for 11 clinically relevant antibiotics, to compare the two techniques. We performed whole genome sequencing analyses on 234 blood culture isolates (genotype) to detect acquired antibiotic resistance. Minimal inhibitory concentrations (MIC) for E. coli were performed for amoxicillin, cefepime, cefotaxime, ceftazidime, meropenem, amoxicillin/clavulanic acid, piperacillin/tazobactam, amikacin, gentamicin, tobramycin, and ciprofloxacin, using the ISO 20776-1 standard broth microdilution method as recommended by EUCAST (phenotype). We then compared the two methods for statistical 'agreement'. A perfect (100%) categorical agreement between genotype and phenotype was observed for gentamicin and meropenem. However, no resistance to meropenem was observed. A high categorical agreement (> 95%) was observed for amoxicillin, cefepime, cefotaxime, ceftazidime, amikacin, and tobramycin. A categorical agreement lower than 95% was observed for amoxicillin/clavulanic acid, piperacillin/tazobactam, and ciprofloxacin. Most discrepancies occurred in isolates with MICs within ± 1 doubling dilution of the breakpoint and 22.73% of the major errors were samples that tested phenotypically susceptible at higher antibiotic exposure and were therefore considered as 'not resistant'. This study shows that WGS can be used as a valuable tool to predict phenotypic resistance against most of the clinically relevant antibiotics used for the treatment of E. coli bloodstream infections.

Serratia marcescens outbreak in a neonatal intensive care unit and the potential of whole-genome sequencing.

BACKGROUND: Serratia marcescens is notorious for its increasing antimicrobial resistance and potential to cause outbreaks in neonatal intensive care units (NICUs). A promising tool in outbreak investigations is whole-genome sequencing (WGS). OBJECTIVES: To describe a S. marcescens outbreak (2018-2019) in an NICU and discuss which infection control measures contributed to containment, addressing the potential of WGS. METHODS: S. marcescens isolates from patients and the environment isolated during the 2018-2019 NICU outbreak were analysed. In comparison, isolates from previous presumed NICU outbreaks and adult blood cultures were included. WGS and whole-genome multi-locus sequence typing analysis were performed. RESULTS: Sixty-three S. marcescens isolates were analysed. The 2018-2019 outbreak was divided into three clusters, including four environmental strains (drains, N=3; baby scale, N=1). The strains differed significantly from those of an NICU outbreak in 2014 and adult blood cultures. Besides standard infection control measures, the siphons were replaced and weekly decontamination was performed with acetic acid 10%. Seven acquired-resistance genes and 29 virulence-associated genes were detected. CONCLUSIONS: It was assumed that both neonates and drains were reservoirs of S. marcescens cross-contamination via the hands of healthcare workers and parents. Initially, standard measures, including hand hygiene, were reinforced. However, definitive containment was achieved only after replacement of the siphons and weekly decontamination with acetic acid. WGS enables faster recognition of an outbreak with accurate mapping of the spread, facilitating the implementation of infection control measures. WGS also provides interesting information about the spread of antibiotic resistance and virulence genes.