In vitro activity of ceftazidime-avibactam, imipenem-relebactam, aztreonam-avibactam, and comparators toward carbapenem-resistant and hypervirulent Klebsiella pneumoniae isolates

ABSTRACT Carbapenem-resistant and hypervirulent Klebsiella pneumoniae (CR-hvKP) strains are increasingly reported, posing a significant threat to public health. Therefore, effective antimicrobial therapy is urgently needed. This study aimed to analyze the in vitro activity of ceftazidime-avibactam (CZA), imipenem-relebactam (IMR), and aztreonam-avibactam (AZA) toward CR-hvKP and carbapenem-resistant Klebsiella pneumoniae (CRKP) isolates. Non-repetitive clinical CRKP and CR-hvKP strains were collected from Sichuan Provincial People’s Hospital between August 2018 and June 2022. CR-hvKP strains were screened using string tests and polymerase chain reaction (PCR). The microbroth dilution method was used to evaluate in vitro antibacterial activity of CZA, IMR, and AZA toward CRKP and CR-hvKP stains. The molecular characteristics of CRKP and CR-hvKP strains were investigated using PCR amplification. The virulence features of CR-hvKP strains were investigated using serum resistance assays and a Galleria mellonella infection model. A total of 114 CRKP and 40 CR-hvKP strains were collected. The susceptibility rates of CRKP and CR-hvKP to tigecycline, colistin, and polymyxin B exceeded 89.5%. The susceptibility rates of CRKP and CR-hvKP to CZA were 64.0% and 77.5%, respectively; the susceptibility rates to IMR were 92.5% and 71.9%, respectively; and the susceptibility rates to AZA were 89.5% and 75.0%, respectively. Multilocus sequence typing and wzi-loci sequencing identified sequence type 11 KL64 as the predominant type in CRKP and CR-hvKP strains. Carbapenemase genes were dominated by blaKPC-2 and blaNDM-1. IMR and AZA may be promising therapeutic agents for the treatment of infections caused by CRKP and CR-hvKP isolates. IMPORTANCE To our knowledge, this is the first study to report the in vitro activity of two novel antimicrobial drugs, including imipenem-relebactam (IMR) and aztreonam-avibactam (AZA), toward carbapenem-resistant and hypervirulent Klebsiella pneumoniae (CR-hvKP) strains. Our in vitro activity study revealed that only few antibacterial agents (including several novel agents) exhibit high antimicrobial activity toward carbapenem-resistant Klebsiella pneumoniae (CRKP) and CR-hvKP isolates. IMR and AZA may be promising therapeutic agents for the treatment of infections caused by CRKP and CR-hvKP isolates.

antimicrobial agents but exhibit greater virulence than cKP, primarily causing commun ity-acquired infections (2,3).Although cKP and hvKP have evolved in a parallel manner over the past few decades, carbapenem-resistant and hypervirulent KP (CR-hvKP) strains are increasingly reported (4)(5)(6).These strains have both antimicrobial resistance determinants and virulence factors on the same plasmids, posing an even greater clinical challenge.
Over the past few years, several novel antimicrobial agents, including plazomi cin, cefiderocol, eravacycline, and β-lactam/β-lactamase inhibitor combinations, such as ceftazidime-avibactam (CZA) and imipenem-relebactam (IMR), were approved by the United States Food and Drug Administration (FDA) for the treatment of compli cated intraperitoneal infections and complicated urinary tract infections (7).Unfortu nately, these β-lactam/β-lactamase inhibitor combinations are primarily used to inhibit serine β-lactamases-producing carbapenem-resistant Enterobacteriaceae (CRE) but not metallo-β-lactamases.In contrast, aztreonam (ATM) is the only monocyclic β-lactam that is not hydrolyzed by metallo-β-lactamases. Aztreonam-avibactam (AZA), another novel β-lactam/β-lactamase inhibitor combination, can, therefore, simultaneously target different carbapenemases, and may be used to treat infections caused by carbapenemresistant Klebsiella pneumoniae (CRKP) and CR-hvKP strains that produce different types of carbapenemases (8).AZA is currently undergoing clinical trials pending regulatory approval (NCT03580044) (8).New antimicrobial agents offer hope for the clinical treatment of CRKP and CR-hvKP infections, which exhibit a concerning epidemic trend in China.However, the antibacterial activity of these new antibacterial drugs toward CR-hvKP in vitro has not been reported.We therefore investigated the in vitro antibacte rial activity of some of these new drugs toward CRKP and CR-hvKP to provide a basis for the clinical treatment of CRKP and CR-hvKP infections.

Bacterial isolates
Between August 2018 and June 2022, a total of 2,529 KP strains were collected from Sichuan Provincial People's Hospital.Among them, 114 CRKP and 40 non-repetitive CR-hvKP isolates were identified.The isolates were obtained from different samples, including sputum, urine, blood, and pus.In this study, CRKP was defined as clinical KP strains that are resistant to at least one carbapenem [ertapenem, imipenem (IPM), meropenem (MEM), or doripenem] according to the 2020 Clinical Laboratory Standards Institute (CLSI) breakpoints(9, 10); hvKP was defined as a positive string test (>5 mm) and co-harboring the genes rmpA/rmpA2 and iucA, as previously described (11).We used PCR and subsequent Sanger sequencing to detect virulence genes.The virulence gene primers are listed in Table S1.

String tests
String tests were performed to identify the hypermucoviscous phenotype as previously described (4).All CRKP and CR-hvKP isolates were inoculated on 5% sheep blood agar plates and incubated at 37°C overnight.The string test was considered positive upon the formation of a viscous string greater than 5 mm in length.

Antibiotic resistance genes and capsular serotyping
We used PCR analyses and subsequent Sanger sequencing to detect carbapenemase genes (blaKPC, blaNDM, blaVIM, blaIMP, blaSME, blaGES, and blaOXA-48-like), as well as other resistance genes (blaCTX-M-1, blaCTX-M-2, blaCTX-M-8, blaCTX-M-9 blaSHV, and blaTEM).The boiling method was used to extract the genomic DNA of KP.The PCR products were visualized on a 1.5% agarose gel.Similarly, PCR analysis was used to detect the capsular type of KP, followed by wzi-loci sequencing, as previously described (14).The specific primers used in this section are listed in Tables S2 and S3.

Serum resistance assay
Serum resistance assays were conducted by mixing CR-hvKP with human serum obtained from 10 healthy volunteers as previously described (4,17,18).Briefly, 100 µL CR-hvKP strains in mid-log phase were mixed to a final concentration of 1 × 10⁶ CR-hvKP CFU/mL with 300 µL human serum in a 1:3 ratio and incubated at 37°C.At 0, 1, 2, and 3 h, 100 µL aliquots were plated on Luria broth agar and incubated at 37°C overnight to determine the number of CFUs.All assays were performed in triplicate.

Galleria mellonella infection models
Among the 40 CR-hvKP strains, 10 strains were randomly selected to verify in vivo virulence in Galleria mellonella.Eight randomly selected G. mellonella larvae, weighing approximately 350 mg (Tianjin Huiyude Biotech Company), were selected for each isolate and maintained on woodchips in the dark at approximately 15°C.The CR-hvKP concentration was adjusted to 1 × 10 7 CFU/mL with PBS, and a 10-µL aliquot was administered via the rear left proleg to infect the G. mellonella, as previously described (4).The survival rates of G. mellonella were recorded every 12 h.The hypervirulent strain SC43-25, isolated from blood samples of a 65-year-old human admitted for liver abscess and subsequent bloodstream infection and endophthalmitis, was selected as the positive virulence control.Whole genome sequence analysis revealed that this strain (GenBank JAVKOU000000000) is highly homologous to the typical hypervirulent strain NTUH-K2044.The classic ST11 strain SC12-22 (GenBank JAVJNH000000000) (string test/iucA/rmpA/rmpA2 negative) was selected as the negative virulence control.These experiments were conducted in duplicate.

Isolate identification
A total of 154 CRKP strains were collected, of which 40 CR-hvKP strains were identified by string tests and PCR amplification of the rmpA/rmpA2 and iucA genes.

Antimicrobial susceptibility
The antimicrobial susceptibility results are summarized in Table 1.In this study, the CRKP and CR-hvKP strains exhibited low (0.0%-14.0%) susceptibility rates to IPM, MEM, CAZ, FEP, CIP, and ATM, and variable susceptibility rates to the novel antibacterial agents.The CR-hvKP group exhibited greater susceptibility rates to IMR (92.5%) than the CRKP group (71.9%).Of the CRKP strains, 89.5% exhibited susceptibility to AZA compared with 75.0% of the CR-hvKP strains.In addition, the susceptibility rates of CRKP and CR-hvKP to CZA were 64.0% and 77.5%, respectively.Both groups exhibited the highest susceptibility rates to COL, POL, and TGC (>89.5%).

Capsular serotyping
The distribution of capsular serotyping in CRKP and CR-hvKP was determined using wzi-typing (Table 3

MLST type
The distribution of MLST typing in CRKP and CR-hvKP is summarized in

DISCUSSION
To our knowledge, this is the first study to report the in vitro activity of two novel antimicrobial drugs, including IMR and AZA, toward CR-hvKP strains.Antimicrobial resistance has become a major global public health challenge, with nearly 5 million infections worldwide in 2019, resulting in at least 1.27 million deaths (19).CRE, most commonly CRKP, plays an important role in the emergence of antimicrobial  resistance, and there are limited treatment options for infections caused by CRKP and CR-hvKP.In our study, all CRKP and CR-hvKP strains exhibited low susceptibility to IPM, MEM, CAZ, FEP, ATM, AMK, and SXT, which is consistent with previous studies (20,21).TGC, COL, and POL exhibited greater antibacterial activity toward both CRKP and CR-hvKP, with susceptibility rates exceeding 89.5%, indicating that they may maintain a high level of activity against these two bacteria.However, the toxic side effects of COL and POL limit their clinical use (22).TGC, a glycylcycline peptide antibacterial drug derived from minocycline, inhibits bacterial protein synthesis and is effective against a wide range of bacteria, including CRKP (23).Its activity is independent of carbapenemase presence and type.However, TGC use is generally limited to the treatment of intraabdominal infections as rapid distribution in tissues after administration limits serum and urinary concentrations.High doses of TGC may be more effective than standard TGC doses for treating complex intra-abdominal infections (24).In recent years, the availability of several novel β-lactam/β-lactamase inhibitor combinations has given hope for the clinical treatment of CRKP.In this study, drug susceptibility tests were performed for IMR, CZA, and AZA.Comparisons between CRKP and CR-hvKP imply that hypervirulence may affect antibiotic susceptibility.However, this could not be validated in this study since the resistance backgrounds of the isolates vary greatly.In China and in Sichuan Provincial People's Hospital, KPC-2-producing ST11 KL64 is the most common type among CR-hvKP strains (27).Therefore, the main putative evolutionary path of CR-hvKP may be a classic KPC-2-producing ST11 KL64 CRKP strain acquiring a virulence plasmid.However, whether this results in a fitness cost, such as a decreased MIC for carbapenems, remains unknown.One study reported by Zhang et al. (27) demonstrated that there was limited fitness cost for three KPC-2-producing ST11 CR-hvKP isolates compared to CRKP.Another study (28) reported that ST23-K1 hvKP isolates might compromise virulence and demonstrate a lower fitness cost.To our knowledge, there is no relevant literature report on the fitness cost of CRKP acquiring a pLVPK-like plasmid.Further studies are needed to evaluate whether acquiring a virulence phenotype influences the antimicrobial susceptibility of a classic CRKP strain.This study has several limitations.Firstly, the identification of CR-hvKP included strains with positive string tests and rmpA/rmpA2 and iucA genes.However, not all hvKP isolates are string-test positive (29), and our results may, therefore, have underestimated the incidence of CR-hvKP.Secondly, delineation of hvKP virulence genes remains incomplete (29), and the combination of virulence genes required for maximal virulence remains unknown.The biomarkers for hypervirulence used in this study (rmpA/rmpA2 and iucA genes), therefore, do not guarantee hypervirulence, and additional animal model studies are necessary.Thirdly, this was a single-center retrospective study, and clinical features and outcomes were not presented.
In conclusion, our in vitro activity study revealed that only few antibacterial agents, including several novel agents, exhibit high antimicrobial activity toward CRKP and CR-hvKP isolates.Among these, two novel β-lactam/β-lactamase inhibitor combina tions-IMR and AZA-may be promising therapeutic agents for treating infections caused by CRKP and CR-hvKP isolates, especially KPC-2-producing ST11 KL64, and further clinical evaluation of the efficacy of these agents is warranted.

FIG 1
FIG1 Virulence potential of CR-hvKP strains in a Galleria mellonella infection model.

TABLE 3
Capsular serotyping of CRKP and CR-hvKP strains

TABLE 4
MLST of CRKP and CR-hvKP strains

TABLE 5
Characteristics of 10 selected CR-hvKP strains