Klebsiella pneumoniae (KP) is a Gram-negative bacterium belonging to the family Enterobacterales and is characterized by its distinct capsule and expression of fimbriae in most strains. Being a member of the normal intestinal and upper respiratory tract flora in humans and animals, KP often causes hospital-acquired infections, such as bacteremia, respiratory tract infection, enteritis, meningitis, wound infection, and urinary tract infection (Wyres et al., 2020). Currently, carbapenem antibiotics that are widely used for treatment in clinical practices in China include imipenem (IPM), meropenem (MEM), and ertapenem (ETM). According to nationwide bacterial resistance surveillance data in China, the resistance rates of KP to MEM and IPM were 2.9% and 3.0% in 2005, respectively, but increased to 26.3% and 25.0% in 2020, respectively. With the continuous increase in drug resistance among pathogens, antibiotic resistance in carbapenem-resistant KP (CRKP) has progressively worsened as resistance rates show certain regional differences (Hu et al., 2020).

Ceftazidime/avibactam (CZA) is a novel cephalosporin/β-lactamase inhibitor that effectively inhibits type A (primarily KPC), type C, and certain type D (OXA-48) carbapenemases, but is ineffective against type B carbapenemases (Falcone et al., 2021; Karaiskos et al., 2021). Currently, antibacterial drug options for the clinical treatment of CRKP are extremely limited, and the development of polymyxin B and tigecycline resistance in patients has been frequently reported (Jin et al., 2021). CZA, a novel antimicrobial agent, possesses good therapeutic effects and safety (Falcone et al., 2021). It also offers other advantages such as high antibacterial activity against multidrug-resistant gram-negative (GN) bacteria, high patient compliance, few adverse effects, and good clinical efficacy. Therefore, CZA is currently widely used in clinical practice as salvage therapy for treatment in patients infected with CRKP (Oueslati et al., 2019).

A new phenotype, identified as hypermucoviscous (hypervirulent) KP (HMKP), was first described in Taiwan in the 1990s and 1980s and mainly caused serious infections in immunocompetent, young, and healthy individuals (Liu et al., 1986; Wang et al., 1998). Of a more serious note, because of the dissemination of mobile genetic elements encoding carbapenemases, carbapenem-resistant HMKP isolates have been increasingly reported (Martin et al., 2017; Peng et al., 2022). The emergence of carbapenem-resistant HMKP isolates is of great concern as they could cause severe untreatable infections in individuals (Di Domenico et al., 2020). Moreover, capsular polysaccharides (CPS) and virulence-associated genes are remarkable virulence factors of KP that allow this bacterium to pose an important threat to public health (Liu et al., 2014). Among the over 78 CPS types of KP, K1 and K2 (and to a lesser extent K5, K20, K54, and K57) strains overproducing capsular polysaccharides are the most virulent human pathogens (Wang et al., 1998; Liu et al., 2014). Of these virulence factors, CPS are considered the major determinant of KP pathogenesis. As determined by multivariate analysis, strains carrying rmpA are significantly associated with the HMV-phenotype (Lin et al., 2004; Yu et al., 2006). Meanwhile, wcaG encodes capsular fucose, which may enhance the ability of the bacteria to evade phagocytosis by macrophages (Fu et al., 2018). The siderophore-associated gene iutA mainly mediates iron transport in the GN bacteria, which are widely spread among HMKP isolates (Compain et al., 2014). At present, many studies have reported HMKP infections, especially pyogenic liver abscesses. The present study aimed to systematically describe the molecular characteristics of carbapenemases and hypermucoviscous traits, as well as clinical features of CRKP isolates. The present study aimed to provide evidence for the investigation and prevention of CRKP infections worldwide. Moreover, our results indicate that available surveillance and strict infection administration strategies should be implemented to prevent carbapenem-resistant HMKP outbreaks in Hangzhou, eastern China.

Currently, CRKP and carbapenem-resistant HMKP isolates, associated with considerable morbidity and mortality, are a major threat to public health worldwide (Wyres et al., 2020). Our study has described the prevalence of CRKP and carbapenem-resistant HMKP isolates from a tertiary hospital in China and characterized the drug resistance, virulence factors, and risk factors of these isolates. Previous studies reported that carbapenemase production was the primary mechanism underlying CRKP drug resistance (Gandra and Burnham, 2020; Karaiskos et al., 2021). CZA, a combination of ceftazidime (cephalosporin) and avibactam (novel synthetic β-lactamase inhibitor), was approved in the Chinese market in May 2019 and serves as an effective option for the treatment of KPC- or OXA-48–producing KP infection (Mehmood et al., 2020; Mataraci et al., 2020). Our results showed that the rates of resistance to tigecycline and polymyxin B in CRKP were 1.2% and 4.8%, respectively, which demonstrates the emergence of drug-resistant isolates at our hospital following the prolonged use of antimicrobials. The rate of resistance to CZA in our studied isolates (26.4%) was lower than that reported in a previous study, which may be attributed to the presence of carbapenem-resistant Enterobacterales isolates producing high levels of NDM (Han et al., 2020). These patterns differ considerably from those observed in Europe. Grundmann et al. (2017) reported that the prevalence rate of OXA-48-producing Enterobacterales was 38% (333/927), which was second to the prevalence rate of KPC-producing Enterobacterales (42%, 393/927), but higher than that of NDM-producing Enterobacterales (12%, 113/927). In the present study, we also observed the absence of CZA-resistant isolates among those that produced KPC-2 or OXA-48 only. Our findings indicate that the carriage of a single gene was the dominant antibiotic genotype carriage pattern in the clinical CRKP isolates from our hospital, which was consistent with the results reported by two previous studies (Bi et al., 2019; Han et al., 2020). In total, 15 of the 417 clinical isolates did not produce carbapenemases; their resistance to carbapenems may be largely related to the combined effects of the absence of membrane porins and other mechanisms (Bulman et al., 2021).

The detection of metallo-β-lactamase-producing carbapenem-resistant Enterobacterales (CRE) strains has limited the clinical applications of CZA, but also promoted the development of novel therapies or combination therapies against drug-resistant CRE strains. For instance, Biagi et al. (2019) reported that aztreonam combined with MEM/vaborbactam or CZA exhibited good antimicrobial activity against NDM-producing CRE strains and non-OXA-48-expressing Enterobacterales. Another report demonstrated that the combination of colistin and amikacin exhibited sustained antimicrobial activity against E. coli co-producing NDM-5 and MCR-1 and may serve as an alternative therapeutic option for the treatment of lethal infections (Zhou et al., 2017).

Furthermore, our results showed that KPC-2, NDM, and OXA-48-like were the most common carbapenemases in the CRKP strains isolated from our hospital, with bla _KPC-2 being the main carbapenemase gene. All CRKP isolates were highly resistant to cephalosporins, carbapenems, and fluoroquinolones, but showed higher susceptibility to polymyxin B and tetracycline. CZA exhibited good antimicrobial effects against clinical CRKP isolates that carried a single bla _KPC-2 or bla _OXA-48-like gene, among which CZA-resistant isolates were absent. The susceptibility rate demonstrated by the present study exceeds the CZA susceptibility of 92.6% previously reported (Kazmierczak et al., 2018). Given the high drug resistance, mortality rate, and anti-infective treatment costs, as well as the difficulty in eradicating CRKP strains from the host, enhanced monitoring of carbapenemase-resistant strains and their prevalence by clinicians is required for the adoption of practical and effective measures to prevent the spread of highly resistant isolates.

Other previous reports suggested that acquiring high virulence and carbapenem resistance in KP was increasingly becoming a concern in hospitals in China (Zhan et al., 2017; Li et al., 2020a). With the increasing aging of the Chinese population, this study showed that such infections are one of the factors that seriously affect health. Hypermucoviscosity and hypervirulence are two different KP phenotypes (Catalán-Nájera et al., 2017). The string test is widely used to identify hypervirulent KP(hvKP). A few classical KP isolates tested positive in the string test, despite being categorized as low virulence isolates. The sensitivity, specificity, and accuracy of the sting test are 95.56, 91.38, and 93.20%, respectively (Li et al., 2020b). Taken together, this study indicated that the string test is still a reliable although imperfect method. A report from Hangzhou, China showed that among 42 carbapenem-resistant HMKP isolates, 16 (38.1%) belonged to K1, 5 (11.9%) belonged to K2, and 11 (26.2%) were non-typable (Chen et al., 2018). Another study from Wenzhou, China found that only one carbapenem-resistant HMKP isolate belonged to K2 (4.5%, 1/22), and that 33.3% (7/21) of the isolates belonged to K20 (Zhan et al., 2017). Similarly, in the present study, K2 was the main capsular serotype and accounted for 46.5% (20/43) of the isolates, followed by K1 (27.9%, 12/43). Moreover, we identified that 25.6% (11/43) of the carbapenem-resistant HMKP isolates belonged to K54, suggesting that K54 may be an important capsular serotype associated with carbapenem-resistant HMKP. A report from China described that rmpA was present in nearly all HMKP isolates (97.6%, 41/42) (Chen et al., 2018). Herein, 39.5% (17/43) of the carbapenem-resistant HMKP isolates harbored rmpA.

Previous studies have shown that two capsule regulator genes, rmpA and rmpA2, are hvKP-specific virulence factors in terms of the abscess formation of HMKP isolates on the virulence plasmid (Li et al., 2020b). For instance, wcaG and magA are associated with capsule biosynthesis (Zhang et al., 2019). Siderophores (SPs) are also essential virulence genes, which can be released from HvKP strains invading the host body. SPs can acquire iron from the binding protein in the host body and can then bind to the receptor to re-enter cells. Four SPs secreted by hvKP, namely salmochelin, enterobactin, yersiniabactin, and aerobactin, have been reported previously (Chen et al., 2018; Han et al., 2022). The function of aerobactin gene iutA is well-correlated with the ex vivo growth/survival of hvKP in humans (Russo et al., 2015). Hence, SP-related encoding genes, especially iutA, are key markers in the identification of hvKP (Russo et al., 2015; Han et al., 2022). Many virulence-based experiments may be required for isolates from hvKP in vitro and in vivo, such as in mice models, greater wax moth (G. mellonella) models, and neutrophil phagocytosis assays. Mouse infection models are widely used to evaluate the virulence of pathogens but have disadvantages, including high expenses and requirement for an animal biosafety level-2 facility environment (Zhang et al., 2019). G. mellonella, a high-throughput animal model developed over the last 20 years, can also be used to determine bacterial pathogenicity (Zhang et al., 2019; Chen et al., 2021). G. mellonella larvae are easily obtained, inexpensive, and can be maintained in a wide temperature range (Zhang et al., 2019). Hence, G. mellonella, is being increasingly used as a model for evaluating the pathogenicity of various microorganisms. This model serves as an alternative system that provides comparable data and is more cost-effective than mammalian testing, ethically acceptable, and highly beneficial. The study adopted the G. mellonella infection model, which can be used to determine bacterial pathogenicity in view of these advantages. Here, we present data indicating that the G. mellonella model is an easily manageable system for comparing the virulence of different K. pneumoniae isolates.

This study evaluated the risk factors associated with clinical characteristics caused by carbapenem-resistant HMKP isolates. Most of the carbapenem-resistant HMKP isolates were associated with significant risk factors such as prior antibiotic therapy, previous hospitalization for 5 days or more, invasive procedures, and mechanical ventilation. Other studies have reported that prolonged hospitalization, invasive procedures (e.g., central or urinary catheter insertion, tracheostomy, and invasive abdominal procedures), serious underlying diseases (e.g., diabetes mellitus, hemiplegia, and solid tumors), and prior carbapenem exposure, were independent predictors for infection development in patients with carbapenem-resistant HMKP isolates (Borer et al., 2012; Chen et al., 2018; Han et al., 2022).

Our study had several limitations. First, multilocus sequence typing and pulsed-field gel electrophoresis clustering were not performed using the CRKP isolates; therefore, the link among the strains could not be divided into different groups. Second, the number of carbapenemase, capsule, and virulence-related genes tested was limited. Third, this is a retrospective study that was conducted in a single hospital.

In conclusion, the resistance rate of CRKP to β-lactam antibiotics was higher than 94%, except for CZA, and it is increasing annually. KPC-2, NDM, and OXA-48-like were the most prevalent carbapenemases among CRKP clinical isolates in China. A small number of the strains carried two or more carbapenemase genotypes. The newly available β-lactam combination agent, CZA, demonstrated good in vitro activity against KPC-2 or OXA-48-like-producing isolates and showed promise in treating CRKP infections. K1 and K2 were the main capsular serotypes of the HMKP strains. Considering that CRKP is highly resistant to commonly used antibiotics, antibiotics for clinical treatment of CRKP infections should be selected following drug sensitivity tests, and hospital administrators should provide training and assessment to clinicians on the rational use of antibiotics to reduce the infection rate of multidrug-resistant bacteria in the future. Our results provide evidence for the investigation and prevention of CRKP infections worldwide.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethics Committee (IRB) of the Second Affiliated Hospital of Zhejiang Chinese Medical University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. The animal study was reviewed and approved by the Ethics Committee (IRB) of the Second Affiliated Hospital of Zhejiang Chinese Medical University. No potentially identifiable human images or data is presented in this study.

XT and DJ conceived and designed the experiments. QW, MC, GM, and JF performed the experiments. QO, LZ, and XC analyzed the data. QW and MC wrote the manuscript. All authors read and approved the final manuscript. All authors contributed toward data analysis, drafting and revising the manuscript, and agreed to be accountable for all aspects of the work. All authors contributed to the article and approved the submitted version.