Next-generation sequencing for typing and detection of resistance genes: performance of a new commercial method during an outbreak of extended-spectrum-beta-lactamase-producing Escherichia coli

Abstract

Next-generation sequencing (NGS) has the potential to provide typing results and detect resistance genes in a single assay, thus guiding timely treatment decisions and allowing rapid tracking of transmission of resistant clones. We evaluated the performance of a new NGS assay (Hospital Acquired Infection BioDetection System; Pathogenica) during an outbreak of sequence type 131 (ST131) Escherichia coli infections in a nursing home in The Netherlands. The assay was performed on 56 extended-spectrum-beta-lactamase (ESBL) E. coli isolates collected during 2 prevalence surveys (March and May 2013). Typing results were compared to those of amplified fragment length polymorphism (AFLP), whereby we visually assessed the agreement of the BioDetection phylogenetic tree with clusters defined by AFLP. A microarray was considered the gold standard for detection of resistance genes. AFLP identified a large cluster of 31 indistinguishable isolates on adjacent departments, indicating clonal spread. The BioDetection phylogenetic tree showed that all isolates of this outbreak cluster were strongly related, while the further arrangement of the tree also largely agreed with other clusters defined by AFLP. The BioDetection assay detected ESBL genes in all but 1 isolate (sensitivity, 98%) but was unable to discriminate between ESBL and non-ESBL TEM and SHV beta-lactamases or to specify CTX-M genes by group. The performance of the hospital-acquired infection (HAI) BioDetection System for typing of E. coli isolates compared well with the results of AFLP. Its performance with larger collections from different locations, and for typing of other species, was not evaluated and needs further study.

FIG 1

Results of amplified fragment length polymorphism (AFLP) analysis. The colors indicate the clusters determined by amplified fragment length polymorphism (AFLP) analysis. Isolates without a color are those considered unique isolates. For the largest cluster (green; 31 isolates), only 5 isolates are presented in the figure for practical reasons. The columns on the right side of the figure indicate (per isolate row) the results of the microarray analysis (resistance genes), the location of isolate collection (NH1 or NH2, followed by the department), and the results of the phylogroup and ST131 PCR analyses. The numbers (in gray boxes) on the far right indicate the WGS isolate number and correspond to the WGS numbers presented in Fig. 2 and 3. NT, not determined.

FIG 2

Phylogenetic tree generated by the HAI BioDetection system in combination with AFLP results. The five AFLP clusters are indicated by colors for each isolate. Isolates considered unique by AFLP are indicated in white. The location of residence (nursing home location NH1 and NH2, building A, B, C, or D, or Department 1, 2, or 3) is indicated in capitals to the right of each circle representing an isolate. Arrows indicate isolates for which whole-genome sequencing was performed. P1, P2, and P3 refer to patients 1, 2, and 3. The scale at the bottom indicates substitutions per site relative to the 10 kb of sequence included in the analysis by the kit.

FIG 3

Phylogenetic tree based on whole-genome sequencing data. (A) Phylogenetic tree based on whole-genome sequencing data from the selected set of isolates (see the text). (B) Phylogenetic tree generated with higher bootstrap support values for a subset of isolates (as indicated in panel A). P1, P2, and P3 refer to patients 1, 2, and 3.