Intensive Care Unit (ICU) is the main department of nosocomial infections and has been considered as a factory for manufacturing, disseminating, and enhancing antimicrobial resistance. The increasing incidence of antimicrobial resistance of isolates from ICU is becoming more and more serious due to the complicated conditions such as extremely vulnerable critically ill patients with many basic diseases, use of invasive operations, relatively long hospitalization time and excessive use of generally broad-spectrum antimicrobials, which bring difficulties to clinical anti-infective treatment[16, 17]. Meanwhile, the infections caused by CRKP are significantly increasing over time, which pose a challenge to clinical treatment and negative impact on patient outcome due to the limited availability of clinically antimicrobials.
In our study, among 133 CRKP strains, 102 were isolated in the intensive care unit, accounting for 76.69%, showing the high prevalence of patients, which was consistence with previous studies[19, 20]. Additional, the isolation rate of CRKP in sputum was the highest (60.90%), which might be due to the heavy use of ventilator reported as the major risk factor of CRKP infections in ICUs.
All the 133 CRKP isolates were found to be carbapenemase-producers and harbor blaKPC-2 gene, which accounted for the most frequent mechanism for resistance to carbapenems in K. pneumoniae. In China, KPC-producing-K.pneumoniae was the most prevailing and reported as the predominant outbreak strain[21, 22]. Fortunately, none multi-carbapenemase-producing K.pneumoniae was observed in this study. However, further analysis revealed the high prevalence of the combination of three β-lactamases genes of blaSHV, blaCTX, and blaTEM among these KPC-2 producing strains. The co-existence of resistant determinants with the combination of carbapenemase and two or more β-lactamases leaded to the multi-drug resistance.
As a result, all the isolates displayed a multi-resistant phenotype, with all resistant to cephalosporins and carbapenems, such as cefazolin, cefuroxime, cefatriaxone, ceftazidime, cefotaxime, cefepime, piperacillin-tazobactam, ampicillin-sulbactam, and imipenem. Although only 3.0% and 2.3% isolates showed resistance to tegafycline and Colistin, respectively, the two antimicrobials had the disadvantages of low plasma concentration and significant toxicity. Thus, treatment of infections caused by these multi-resistant isolates are extremely challenging due to its limited antimicrobials and high levels of morbidity and mortality.
According to MLST typing, ST11 was the most dominant sequence type of the 50 CRKP isolates, accounting for 90.0%, which was consistent with previous studies in China[15, 19]. In addition, ST11 clone is the single locus variant of ST258, and that they belong to the same clonal members of CG258, which has contributed largely to the global dissemination of KPC carbapenemase over the past 20 years[24, 25]. The difference is that the ST258 clone is prevailing in North America and Europe, while the ST11clone is frequently isolated in Asia. Moreover, ST11 clone showed multidrug-resistant phenotype with high carriage rate of virulence genes. These virulence factors are conducive to binding, biofilm formation, colonization, and anti-phagocytosis, enhancing environmental survival and, which can successfully demonstrate their persistence and quick expansion across the world[26, 27] .
In addition to ST11, five ST15 KPC-2-producing K. pneumoniae were isolated in patients in neurosurgical ICU, gastroenterology ward, emergency ICU and outpatient, relatively. It was worth noting that K. pneumoniae isolates of ST15 was widely disseminated in Europe and repeatedly reported to spread carbapenemase genes such as NDM-1 in North African and in Greece and OXA-48 in Spain. However, ST15 is gradually detected in China[31, 32]. Moreover, in our study one ST15 strain was isolated from one outpatient who had not been hospitalized or received health care services in the previous 3 months, suggesting that ST15 is a potential risk clone to be a community onset which deserves routine surveillance to prevent outbreak.
MLST is appropriate for genotyping to further provide data not only for local epidemiological investigations but also for global and evolutionary studies. However, since MLST only involve seven housekeeping genes, it is less discriminating than PFGE to analyze outbreaks in different wards. In our study, all the ST11 isolates were divided into A-G seven different PFGE clusters and related sub-clusters. Of which, sub-cluster A1 and G1 appeared the two most predominant clones with > 95% similarity of PFGE patterns. And the sub-cluster A1 consisted of all the isolates collected in emergency ICU, expect for two isolates in neurosurgical, and sub-cluster G1 in both neurosurgical ICU and emergency ICU, suggesting that the two prevalent clones were transmitted in the two ICUs. In addition, there were two isolates assigned to cluster B and one isolate belonged to cluster F were also detected in emergency ICU, which suggested us that emergency ICU was one of the most high-risk wards and might be the focal point of active monitoring for infection control.
Finally, the six isolates belonged to cluster D and E were observed just in cardiothoracic surgery ICU, except one in an outpatient, which might had different genetic backgrounds from the isolates in neurosurgical ICU and emergency ICU. It was worth noting that the CRKP isolates assigned to cluster A, B, C, E and G all had the ability to cause blood infection, which deserved our attention.
In conclusion, our study revealed that ICUs were the main wards for rapid and wide spread of CRKP and clonal dissemination and outbreak of KPC-2-producing CRKP belonged to ST11 was observed in ICUs, especially in emergency ICU in our hospital. Hence,routine surveillance and effective infection control strategies are urgently warranted on these high-risk international strains to prevent outbreak and nosocomial-acquired infection.