Covert dissemination of carbapenemase-producing Klebsiella pneumoniae (KPC) in a successfully controlled outbreak: long- and short-read whole-genome sequencing demonstrate multiple genetic modes of transmission

Abstract

Background: Carbapenemase-producing Enterobacteriaceae (CPE), including KPC-producing Klebsiella pneumoniae (KPC-Kpn), are an increasing threat to patient safety.

Objectives: To use WGS to investigate the extent and complexity of carbapenemase gene dissemination in a controlled KPC outbreak.

Materials and methods: Enterobacteriaceae with reduced ertapenem susceptibility recovered from rectal screening swabs/clinical samples, during a 3 month KPC outbreak (2013-14), were investigated for carbapenemase production, antimicrobial susceptibility, variable-number-tandem-repeat profile and WGS [short-read (Illumina), long-read (MinION)]. Short-read sequences were used for MLST and plasmid/Tn4401 fingerprinting, and long-read sequence assemblies for plasmid identification. Phylogenetic analysis used IQTree followed by ClonalFrameML, and outbreak transmission dynamics were inferred using SCOTTI.

Results: Twenty patients harboured KPC-positive isolates (6 infected, 14 colonized), and 23 distinct KPC-producing Enterobacteriaceae were identified. Four distinct KPC plasmids were characterized but of 20 KPC-Kpn (from six STs), 17 isolates shared a single pKpQIL-D2 KPC plasmid. All isolates had an identical transposon (Tn4401a), except one KPC-Kpn (ST661) with a single nucleotide variant. A sporadic case of KPC-Kpn (ST491) with Tn4401a-carrying pKpQIL-D2 plasmid was identified 10 months before the outbreak. This plasmid was later seen in two other species and other KPC-Kpn (ST14,ST661) including clonal spread of KPC-Kpn (ST661) from a symptomatic case to nine ward contacts.

Conclusions: WGS of outbreak KPC isolates demonstrated blaKPC dissemination via horizontal transposition (Tn4401a), plasmid spread (pKpQIL-D2) and clonal spread (K. pneumoniae ST661). Despite rapid outbreak control, considerable dissemination of blaKPC still occurred among K. pneumoniae and other Enterobacteriaceae, emphasizing its high transmission potential and the need for enhanced control efforts.

Figure 1.

Patient admissions to LTH liver unit and other LTH wards during the outbreak period. Red vertical marks are the time of first KPC detection in the patients, black vertical marks denote the starts and ends of the infection risk periods defined for SCOTTI, green vertical marks denote the closest KPC-negative screening time points (where applicable) before or after the first KPC-positive screening result. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 2.

Phenotypic sensitivity and antimicrobial resistance gene prediction results of 23 strains from LTH KPC outbreak (MICs: mg/L measured by agar dilution at AMRHAI). AMC, co-amoxiclav; FOX, cefoxitin; CIP, ciprofloxacin; TZP, piperacillin/tazobactam; MEM, meropenem; IPM, imipenem; ETP, ertapenem; GEN, gentamicin; CAZ, ceftazidime; TOB, tobramycin; COL, colistin; AMK, amikacin; TEM, temocillin; TGC, tigecycline. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 3.

Outbreak KPC plasmids and contigs derived from long-read sequencing of isolates and their alignment to the most closely genetically matched complete plasmid sequences from NCBI. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 4.

Transmission analysis results inferred by SCOTTI for the dominant outbreak strain KPC-Kpne ST661 using consensus chromosomal sequences and epidemiological data. The ‘maximum clade credibility’ of the posterior coalescent tree (scale in years) was inferred by SCOTTI, taking into account epidemiological data and transmission processes. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.