A Polyclonal Spread Emerged: Characteristics of Carbapenem-Resistant Klebsiella pneumoniae Isolates from the Intensive Care Unit in a Chinese Tertiary Hospital

Klebsiella pneumoniae is regarded as an opportunistic Gram-negative pathogen that can cause several infections such as pneumonia, urinary tract infections, and bloodstream infections (Magill et al. 2014). Due to the overuse of carbapenems for treating severe infections caused by extended-spectrum β-lactamases (ESBLs)-producing bacteria, carbapenem-resistant K. pneumoniae (CRKP) has rapidly increased globally in the past decade (Logan and Weinstein 2017). The expression of plasmid-mediated carbapenemases has been the primary mechanism of carbapenem resistance, and K. pneumoniae carbapenemase (KPC) is the most frequent type found in this species (Martin and Bachman 2018). Among these isolates, ST11 was the predominant clone responsible for disseminating the resistance gene bla_KPC-2 in China (Qi et al. 2011), whereas ST258 accounted for the large majority of KPC-producing K. pneumoniae in the world (Kitchel et al. 2009; Hammerum et al. 2010). Moreover, additional types of carbapenemases have also emerged in K. pneumoniae like NDM-1 and OXA-48 categorized as class B metallo-β-lactamase (MBL) and class D enzymes, respectively, which confer specific levels of resistance to carbapenems. Ever since NDM-1 was discovered in K. pneumoniae isolate collected from a Swedish patient who had been hospitalized in India in 2008 (Yong et al. 2009), twenty-four NDM variants have been identified. It poses a significant threat to public health and a severe challenge for clinical treatments (Wu et al. 2019).

ICU hospitalization itself has been considered as an independent risk factor for CRKP acquisition (Schwaber et al. 2008; Hussein et al. 2009; Debby et al. 2012). The estimated detection rate of CRKP in patients admitted to intensive care units increased by 75% in a 20-year surveillance study in China (Tian et al. 2019). The gastrointestinal carriage rate of CRKP among ICU patients could reach 39.0–74.5%. It can be recognized as a reservoir of CRKP for progression from colonization to infection and the potential route of transmission of carbapenem resistance genes (Bratu et al. 2005; Snitkin et al. 2012; Papadimitriou-Olivgeris et al. 2013). Additionally, ICU is often deemed the epicenter of nosocomial infections caused by multidrug-resistant organisms (MDRO) due to the burdens of the vulnerable populations of critically immunocompromised patients and multiple invasive procedures. Thus, the outcome of patients with CRKP infections is inferior, leading to higher mortality in the setting of ICU associated with limited therapeutic options (Vardakas et al. 2015).

Herein, we report an investigation of CRKP carriage and acquisition in the ICU to illustrate the clonal spread of CRKP isolates, their phenotypic and genotypic characteristics, and to track their evolutionary traits further.

Demographic and clinical characteristics of CRKP carriers. A total of 30 CRKP strains were isolated from 26 patients in the intensive care unit (ICU), and the strains from different isolation sites of the same patient were included in this study. Among these CRKP isolates, 12 (40%) were isolated from sputum specimens, and the remaining isolates were obtained from other types of specimens including blood (three, 10%), wound (two, 6.7%), drainage fluid (four, 13.3%), urine or urinary catheter (seven, 23.3%), and bronchial perfusate (one, 3.3%). 69.2% (18) of the patients were male, and 76.9% (20) were over 60 years old.

All patients had undergone invasive procedures such as tracheal intubation and central venous catheterization. During the ICU admission, multiple antimicrobials were used for the treatment of various intercurrent infections. Among the 26 patients with CRKP acquisition, five died, six declined further therapy, and 15 were discharged from the hospital ward. The times from acquisition of CRKP to death for the five patients who died, listed in order, were four, 107, 29, 36, and 16 days, respectively. Other detailed records of patients and information on the bacteria were summarized in Table I.

Antimicrobial susceptibility. The isolates in this study were resistant to nearly all clinically available antimicrobials; more than half of the isolates were only susceptible to one or two kinds of antimicrobials and were called extensive drug-resistant isolates. All isolates presented resistance to ertapenem, amoxicillin/clavulanic acid, cefoperazone/sulbactam cefazolin, cefoxitin, and ampicillin. The isolates were relatively susceptible to four antimicrobials: gentamicin, tigecycline, tobramycin, and amikacin, to which the resistance rates were 36.7, 30.0, 26.7, and 16.7%, respectively. The percentage of resistant isolates to each antibiotic was shown in Fig. 1.

Fig. 1.

Profiling of resistance determinants. The majority of 30 CRKP (n = 23, 76.7%) isolates were positive for the mCIM test. The results of PCR and DNA sequencing showed that nine isolates (30%) harbored the bla_NDM-5 gene, 12 isolates 40% harbored the bla_KPC-2 gene, and one isolate had the bla_NDM-1 gene. The coexistence of bla_NDM-1 and bla_KPC-2 in one isolate was also noticed. The carbapenemase-encoding genes were not detectable in seven isolates.

Bacterial clonal relatedness. Among 30 CRKP isolates, six sequence types (STs) were detected, namely ST722, ST1446, ST111, ST896, ST290, and ST11 as shown by MLST. ST290 and ST11 accounted for 20% (6/30) and 56.7% (17/30) of all isolates tested, whereas the other STs were sporadic. Three ST11 isolates carried bla_NDM-5, yet bla_KPC-2 was more likely to be identified among ST11 clones. By contrast, ST290 clones harbored only bla_NDM-5. Seven isolates that were negative for mCIM test belonged to diverse STs (ST722, ST1446, ST290, ST11, and ST111). Notably, two K. pneumoniae NDM-producers failed to be classified into any sequence types and there was the sole isolate that was negative for either mCIM or sequence typing. Figure 2 displays the annotated minimum spanning tree showing that ST1446, ST111, ST896, ST290, and ST11 clones belonged to different groups, except for ST722 listed in the sub-group of ST11 group.

Fig. 2.

In 2017, WHO published a global priority pathogens list of antibiotic-resistant bacteria, in which CRE was ranked among the Priority 1 pathogens (WHO 2017). Carbapenem-resistant K. pneumoniae, which is the most common carbapenem-resistant Enterobacteriaceae (CRE), has already generated a worrisome crisis of epidemiological, clinical, and infection control issues worldwide (Bradford et al. 2004; Maltezou et al. 2009), including the ICUs across China (Zhang et al. 2011; Yu et al. 2012; Yu et al. 2019).

In this study, we have described the polyclonal spread of CRKP isolates of six distinct sequence types (ST290, ST11, ST722, ST1446, ST111 and ST896) in the same ward (ICU). Among these isolates, two novel clones of ST722 and ST1446 were found, and they did not produce carbapenemases since the negative results of the mCIM test were obtained. It inferred that other mechanisms of resistance might be relevant such as hyperproduction of ESBL enzymes, AmpC β-lactamases, or alteration of outer membrane porins as well as regulation of efflux systems (Kaczmarek et al. 2006; Bush and Jacoby 2010; Filgona et al. 2015). The minimum spanning tree demonstrated that ST722 clone probably shared the same ancestor with ST11 clone in the evolutionary process. As for ST111, it has long been identified among carbapenem-resistant, and ESBL-producing K. pneumoniae isolates, e.g., obtained from Riyadh (uz Zaman et al. 2014), South India (Kumar et al. 2018), New York (Diago-Navarro et al. 2014), and New Zealand (Lester et al. 2011). Further, the ST11 K. pneumoniae, one type of clone responsible for the outbreak of multi-drug carbapenem-resistant K. pneumoniae in Riyadh, carried the OXA-48 gene, suggesting that the acquisition of carbapenem-resistance genes by K. pneumoniae of different STs may contribute to the emergence of diverse CRKP clones. By contrast, only one ST896 CRKP isolate that was identified from Heilongjiang Province in China harbored the bla_IMP-4, bla_SHV, and bla_TEM genes (Gong et al. 2018).

Therefore, this is also the first report on the observation of the bla_KPC-2-harboring ST896 CRKP clone in China.

ST290 CRKP isolates possessing the bla_NDM-5 gene were found disseminated in the ICU in this study. Interestingly, an outbreak of ST290 CRKP with blaNDM-5-positive took place in the same hospital’s wound ward, as we have reported previously (Wang et al. 2019), raising speculation that intra-hospital transmission might be one of the contributory factors to this situation. However, in line with previous reports (Qi et al. 2011; Hu et al. 2016), ST11 was still the predominant type of CRKP clone, accounting for more than half of the isolates. Of note, three CRKP isolates of ST11 clone harbored bla_NDM-5 instead of bla_KPC-2, showing that the genotypic shift occurred in ST11 clone, which was probably attributed to the free movement of plasmidborne genes responsible for carbapenem resistance among the clones within species as stated previously (Mathers et al. 2015). Moreover, the rise in the number of NDM-5-positive isolates of the ST11 clone generates the awareness of the propagation of the MBL genes in high-risk K. pneumoniae clones. ST11 clone could be a suitable vector for the rapid spread of carbapenem resistance mediated by genetic components such as plasmids, integrons, or transposons.

The ICU patients in this study suffered from at least three underlying diseases and underwent several surgical procedures causing damage of mucosal barriers, which could increase the risk of CRKP colonization and infection (Kofteridis et al. 2014). We found that 71.4% of patients aged over 80 years died, as did 85.7% of the patients who stayed in the ICU for more than one month, reflecting the challenges in managing the comorbidities of ICU patients. Additionally, the prolonged hospitalization of CRKP carriers may increase the frequency of patient-to-patient transmission of antimicrobial resistance. According to the medical records, local empirical antibiotic therapy could be administered in the ICU, e.g., cefepime combined with amoxicillin/clavulanic acid (Ji et al. 2015). Ceftazidime/avibactam, a relatively new salvage therapy against CRKP, was also used in some patients with different outcomes (those who survived, died, or for whom the treatment was discontinued). However, the resistance to these antibiotics has previously been revealed due to the MBL production, KPC-2 point mutation, and high KPC expression (Zhang et al. 2019). It indicates that the treatment with ceftazidime/avibactam against MBL-producing CRKP might fail if local variations in epidemiology and genomic evolution of antimicrobial resistance are not tracked. Hence, the focus on characteristics of CRKP in the ICU not only plays an essential role in guiding clinical practice for antibiotic use but also provides the recent information about the evolution of antimicrobial resistance and helps to assign urgently needed tactics for combating the spread of CRKP clones.

The limitations of this study are that we did not distinguish colonization from infection with CRKP, and the small number of isolates was insufficient to illustrate the prevalence of CRKP in the ICU comprehensively.

In conclusion, this is the first report on the polyclonal emergence of six unique STs (ST722, ST1446, ST111, ST896, ST290, and ST11) found in the same ward and on two ST722 and ST1446 clones as being the novel STs in China. Pathogens in the ICU evolve all the time due to intra- and inter-species interactions by the horizontal transfer of antibiotic resistance genes. The emergence of sporadic clones producing MBLs, e.g., producing NDM-5 CRKP isolates of ST290 and ST11 clones, is a warning signal of the genotypic switch in epidemic KPC-producing CRKP population. Therefore, valid interventions should be developed to avoid outbreaks caused by novel subclones.