Epidemiological and genetic characteristics of clinical carbapenem-resistant Acinetobacter baumannii strains collected countrywide from hospital intensive care units (ICUs) in China

Abstract

Acinetobacter baumannii is one of the key Gram-negative pathogens that can cause serious nosocomial infections. In China, a large proportion of clinical A. baumannii strains are multidrug resistant, among which strains resistant to carbapenems are particularly worrisome, as infections caused by such strains may limit the choice of existing antibiotics. We conducted a nationwide and genome-based surveillance on the prevalence and antibiotic susceptibility profile of carbapenem-resistant A. baumannii (CRAB) strains collected from intensive care units (ICUs) in hospitals in different provinces and investigated the routes of transmission and mechanism of resistance by whole-genome sequencing and phylogenetic analysis. We found that CRAB strains were prevalent in 71.4% (55/77) of the ICUs surveyed. Clonal spread of CRAB was found in 37.6% (29/77) of ICUs and a total of 22 different clones were identified. Most clones were transmissible within one ICU, but up to six clones could be detected in at least three hospitals. In addition, carbapenem-hydrolysing class D β-lactamases (CHDL) were found to be mainly responsible for carbapenem-resistance in A. baumannii and the ST2 global-clone is the predominant type of CRAB in China. Importantly, we found that CRAB isolates currently exhibited an extremely low rate of resistance to colistin (0.4%) and tigecycline (2.5%), but a high rate of resistance to ceftazidime-avibactam (70.2%). Findings in this work shall facilitate development of appropriate antimicrobial regimens for treatment of CRAB infections. Further surveillance and research on the evolutionary and epidemiological features of clinical CRAB strains are necessary.

Keywords: Acinetobacter baumannii; OXA-23; carbapenem resistance; clonal transmission; molecular epidemiology.

Figure 1.

Prevalence of CRAB strains in different provinces of China. (a) The prevalence of CRAB strains recovered from the faecal swabs of patients in different provinces of China. (b) The prevalence of CRAB strains recovered from the respiratory samples of patients in different provinces of China. Different shades of colour represent different prevalence levels of carbapenem resistance.

Figure 2.

Antimicrobial susceptibility of 245 CRAB strains isolated from ICUs. IMP, imipenem; MEM, meropenem; CAZ, ceftazidime; SXT, trimethoprim/sulfamethoxazole; TZP, piperacillin/tazobactam; CAV, ceftazidime–avibactam; PIP, piperacillin; FEP, Cefepime; CHL, chloramphenicol; CST, colistin: TGC, tigecycline; CIP, ciprofloxacin; GEN, gentamicin: AMK, amikacin; TET, tetracycline; SCF, cefoperazone–sulbactam. Interpretations of resistance phenotypes follow those of the CLSI (Clinical and Laboratory Standards Institute) M100-S26 document.

Figure 3.

Phylogenetic tree of 245 CRAB strains collected from ICUs in various parts of China during the period July to September 2020. Circle 1 depicts strains isolated from various provinces or municipal cities in China; Circle 2 denotes distribution of MLST types. Strains in each clade are depicted in the same colour and are regarded as clonally disseminated. Circle 3 depicts 5 ICUs in which four or more clones were recovered.

Figure 4.

Minimum spanning tree constructed on the basis of cgMLST allelic genes of 245 CRAB strains isolated from ICUs in China. Each circle depicts an allelic profile based on sequence analysis of 2050 cgMLST genes. The length of the connecting lines represents the number of target genes with different alleles. (a) Colours of the circles denote different cgMLST types. Closely related genotypes (less than 10 alleles difference) are shaded in same node, and clusters are numbered consecutively (1–40). Red arrows indicate that strains belonged to ST1555, the remaining strains are all identified as ST2 type (b) Colours of the circles represent different isolation sources.

Figure 5.

AMRs gene analysis of 245 CRAB strains collected nationwide in China. Heatmaps were obtained by aligning the draft genome sequence of each strain to the AMR gene database. Acinetobacter baumannii are clustered using a maximum likelihood tree. Red and light cyan blue depict the presence and absence of AMRs in the test strains, respectively. The right side denotes the category of AMRs. The colour bar above depicts the geographical distribution of CRAB isolates, with each region being shown by a specific colour in the dataset legend.

Figure 6.

Genetic environment of acquired bla_OXA–23 and bla_KPC-2 elements in CRAB isolates created by Easyfigure. (a) Structure alignment of Tn2006 and Tn2009. Red depicts bla_OXA-23 genes; Green and yellow denote putative functional protein and insertion sequence, respectively. (b) Linear alignment of chromosomal fragments MS114413-chr, R365-1 matching Illumina contigs and plasmid pKP20194a-p2. Mobile elements are highlighted in yellow and drug-resistance gene are depicted by red arrows. Blue frame depicts class I integron. Plasmid backbone of pKP20194a-p2 is highlighted in grey.