Comparative analysis of IR-Biotyper, MLST, cgMLST, and WGS for clustering of vancomycin-resistant Enterococcus faecium in a neonatal intensive care unit

Abstract

Healthcare-associated infections caused by vancomycin-resistant Enterococcus faecium (VREFM) pose a significant threat to healthcare. Confirming the relatedness of the bacterial isolates from different patients is challenging. We aimed to assess the efficacy of IR-Biotyper, multilocus sequencing typing (MLST), and core-genome MLST (cgMLST) in comparison with whole-genome sequencing (WGS) for outbreak confirmation in the neonatal intensive care unit (NICU). Twenty VREFM isolates from four neonates and ten control isolates from unrelated patients were analyzed. Genomic DNA extraction, MLST, cgMLST, and WGS were performed. An IR-Biotyper was used with colonies obtained after 24 h of incubation on tryptic soy agar supplemented with 5% sheep blood. The optimal clustering cutoff for the IR-Biotyper was determined by comparing the results with WGS. Clustering concordance was assessed using the adjusted Rand and Wallace indices. MLST and cgMLST identified sequence types (ST) and complex types (CT), revealing suspected outbreak isolates with a predominance of ST17 and CT6553, were confirmed by WGS. For the IR-Biotyper, the proposed optimal clustering cut-off range was 0.106-0.111. Despite lower within-run precision, of the IR-Biotyper, the clustering concordance with WGS was favorable, meeting the criteria for real-time screening. This study confirmed a nosocomial outbreak of VREFM in the NICU using an IR-Biotyper, showing promising results compared to MLST. Although within-run precision requires improvement, the IR-Biotyper demonstrated high discriminatory power and clustering concordance with WGS. These findings suggest its potential as a real-time screening tool for the detection of VREFM-related nosocomial outbreaks.

Importance: In this study, we evaluated the performance of the IR-Biotyper in detecting nosocomial outbreaks caused by vancomycin-resistant Enterococcus faecium, comparing it with MLST, cgMLST, and WGS. We proposed a cutoff that showed the highest concordance compared to WGS and assessed the within-run precision of the IR-Biotyper by evaluating the consistency in genetically identical strain when repeated in the same run.

Keywords: IR-Biotyper; VREFM; nosocomial outbreak.

Fig 1

The phylogenetic tree and clustering results obtained from the analysis of WGS data using pairwise comparison using de novo references (PCDR). The 20 isolates from 4 patients in the neonatal intensive care unit marked with an asterisk.

Fig 2

The dendrogram obtained using the IR-Biotyper. The blue vertical dashed lines represent clustering cut-off values (COV). The 20 isolates from 4 patients in the neonatal intensive care unit marked with an asterisk. The red box indicates the two outlier isolates identified in the within-run precision study.

Fig 3

Adjusted Rand index (ARI) according to clustering cut-off for IR-Biotyper. The red vertical dashed line represents the optimal clustering cutoff range with the highest ARI (n = 30).

Fig 4

Relationship between methods for clustering (n = 30). Abbreviations: PICU, pediatric intensive care unit; NICU, neonatal intensive care unit; CICU, cardiac intensive care unit; EFM, Enterococcus faecium; ST, sequence type; CT, complex type.