Klebsiella pneumoniae is a common Gram-negative opportunistic pathogen causing gastrointestinal, urinary tract, respiratory and blood-stream infections in clinical practice (Xu et al., 2017; Wyres et al., 2020). Carbapenems are known as ‘last-resort’ β-lactam antimicrobial agents for controlling and treating K. pneumoniae infections (Queenan and Bush, 2007). However, the detection rates of carbapenem-resistant K. pneumoniae (CRKP) have shown an increasing trend annually in recent years (Holt et al., 2015; Wyres et al., 2020). The production of carbapenemases is the primary mechanism mediating carbapenem resistance. Carbapenemases are typically categorized into three major classes: K. pneumoniae carbapenemase (KPC) (Ambler class A with serine-based hydrolytic activities), metallo-β-lactamases (MβL) (Ambler class B with zinc in their active sites, e.g., NDM), and OXA-48 (Ambler class D). KPC carbapenemase is the most prevalent class worldwide, especially KPC-2 and KPC-3 (Ferrari et al., 2019; Wyres et al., 2020; Li et al., 2022; Zhang et al., 2023).

The high incidence of CPKP infections significantly increases the healthcare costs, as well as the morbidity and mortality rates of patients (Wyres et al., 2020). A prospective, multicenter, cohort study showed that the unadjusted 30-day mortality rate due to CRKP infection stood at 12–28% globally (Wang et al., 2022). Besides, a systematic review and meta-analysis indicated that patients infected with CRKP have higher pooled mortality (42.14%) compared to those with carbapenem-susceptible K. pneumoniae (CSKP) infections (21.16%) (Xu et al., 2017). Effectively treating the CRKP infections presents a major challenge to global health. The options for antimicrobial treatment against CRKP infections are limited, with only a few antibiotics are available, including polymyxins, tigecycline, aminoglycosides, fosfomycin and ceftazidime/avibactam (Livermore, 2005; Falagas et al., 2016; Shi et al., 2020; Mohapatra et al., 2021). However, resistance to these agents has been increasingly reported (Gu et al., 2016; Zhang et al., 2017; Lv et al., 2020). Among them, ceftazidime/avibactam, a novel β-lactam-plus-β-lactamase inhibitor combination, is known to be effective against serine β-lactamases but not against metallo-β-lactamases (Shi et al., 2020). With the widespread use of ceftazidime/avibactam, more than 150 KPC-2/KPC-3 variants (hereinafter referred to as KPC variants) that confer resistance to ceftazidime/avibactam have been reported in clinics worldwide, with the majority of these variants discovered in the past 3 years (Mueller et al., 2019; Wu et al., 2022). Compared to KPC-2/KPC-3, the new variants have undergone insertions, deletions, and/or point mutations, altering the KPC structure and enhancing the affinity for ceftazidime while weakening the affinity for avibactam (Mueller et al., 2019; Wu et al., 2022). In the face of rising antibiotic resistance, treating bacterial infections are increasingly difficult.

In the current era, characterized by a low success rate in antibiotic drug discovery, combining two or more antimicrobial agents to treat a single infection, referred to as antimicrobial agent combinations, may present a viable alternative for treating CRKP infections. The European Society of Clinical Microbiology and Infectious Diseases (ESCMID) guidelines propose that, for patients with severe infections caused by carbapenem-resistant Enterobacterales (CRE) strains that exhibit in vitro susceptibility only to polymyxins, aminoglycosides, tigecycline, or fosfomycin, or in the case where β-lactam/β-lactamase inhibitor combinations are unavailable, it is advisable to pursue a therapeutic combination regimen active in vitro (Paul et al., 2022). Antimicrobial agent combinations can exploit four principal effects on microbes: additive effects, where the impact is the same as the combined effects of the individual antibiotics; synergistic effects, where the impact is greater than the sum of the individual antibiotics’ effects; antagonistic effects, where the impact is reduced compared to the combined effects of the individual antibiotics; and indifferent effects, where one antimicrobial agent can’t affect the activity of the other. Most antimicrobial agent combinations produce a synergistic effect and are more potent than either monotherapy (Torella et al., 2010). Indeed, a noteworthy 30% of the novel antibiotics sanctioned by the FDA during the preceding five-year period have been combination therapies (Geraldine et al., 2020). A systematic review, which analyzed 136 studies, revealed that combinations based on colistin and/or carbapenems demonstrated a notable synergistic effect on CRKP strains. This combination significantly enhanced bactericidal activity while concurrently reducing the emergence of antibiotic-resistant bacteria (Scudeller et al., 2021). Notably, a study found that the effectiveness of different antimicrobial agent combinations was influenced by CRKP infections with different resistance mechanisms (Liu et al., 2021). Wu et al. also reported the ceftazidime/avibactam-meropenem combination was highly synergistic against KPC-producing isolates (91.3%) and carbapenemase-non-producing isolates (100%) (Yun et al., 2024). Moreover, the therapeutic strategy of some antimicrobial combinations can also minimize side effects associated with medication dosage for the patient, as it allows for the use of reduced concentrations of each drug, especially in cases where long-term use of antibiotics causes side effects (Frank et al., 2022).

So far, a large amount of researches on the emergence of carbapenem resistance among K. pneumoniae are widely available. Nevertheless, results on the valid antimicrobial agent combinations for the treatment for CRKP infections, particularly for those producing the KPC variant carbapenemases, are limited. In this study, we conducted a comprehensive analysis to explore the in vitro efficacy of 24 antimicrobial agent combinations based on meropenem, polymyxin, tigecycline, ceftazidime/avibactam, respectively, against 44 CRKP strains that encoded carbapenemases, especially those producing the KPC variant of carbapenemases.

The increasing spread of multidrug-resistant (MDR) CRKP strains has further limited therapeutic options for nosocomial infections, resulting in high fatality rates among patients with compromised immunity (Holt et al., 2015). The current therapeutic challenges in treating CRKP infections highlight the urgent need for effective antimicrobial combinations. While numerous studies have explored various antimicrobial agent combinations against CRKP, there is a lack of comprehensive research on therapeutic strategies for CRKPs that produce different carbapenemases, particularly those producing KPC variants. In this study, we identified the in vitro efficacy of 24 antimicrobial combinations with different modes of action in treating of 44 carbapenemase-producing CRKP strains. The combination therapy of antimicrobial agents with different mechanisms of action and pharmacokinetic properties can enhance antibacterial efficacy and reduce the emergence of antibiotic resistance.

In this study, the three most effective antimicrobial combinations against CRKP strains were all based on polymyxin, including polymyxin-aztreonam, polymyxin-meropenem and polymyxin-levofloxacin. Polymyxin, a cationic polypeptide, binds to negatively charged lipopolysaccharides (LPS) and disrupts the bacteria’s outer membrane, thereby exerting its antibacterial effect (Mohapatra et al., 2021). Aztreonam and meropenem function by inhibiting the biosynthesis of bacteria cell walls (Hellinger and Brewer, 1999). Levofloxacin, on the other hand, blocks DNA replication, leading to bacterial cell death (Xue et al., 2021). Consequently, these combinations have demonstrated remarkable therapeutic efficacy against CRKP strains through distinct mechanisms. Additionally, the studies by Yu et al. have also observed synergistic effects of polymyxin-meropenem and polymyxin-levofloxacin against polymyxin-resistant CRKP isolates (Yu et al., 2019).

Reports indicate that polymyxin resistance can develop within 24 h of polymyxin monotherapy, but this resistance is mitigated and delayed when combination therapy is used (Le Minh et al., 2015). Significantly, a retrospective cohort study demonstrated that combination therapy with polymyxin-meropenem improved survival rates in patients suffering from bacteremia caused by KPC-producing K. pneumoniae, in comparison to patients who received polymyxin monotherapy (Le Minh et al., 2015).

KPC-2 is the most prevalent carbapenemase among CRKP strains in China (Zhang et al., 2017). Ceftazidime/avibactam stands as one of the few effective treatment alternatives against KPC-2-producing CRKP strains. But the clinical use of ceftazidime/avibactam has led to mutations in KPC-2 carbapenemases, emerging as a major mechanisms behind ceftazidime/avibactam resistance (Shi et al., 2020). To date, limited studies have investigated antimicrobial combinations targeting KPC variant-producing CRKP strains. In our study, ten antimicrobial agent combinations exhibited synergistic or addictive effects on ≥50% of KPC variant-producing CRKP strains, while only three combinations showed addictive effects on ≥50% of KPC-2-producing CRKP strains.

In addition to the three antimicrobial agent combinations of polymyxin-aztreonam, polymyxin-meropenem, and polymyxin-levofloxacin, which are beneficial for treating KPC variant-producing and KPC-2-producing CRKP strains, the combination of ceftazidime/avibactam with one other active drug (such as aztreonam, meropenem, imipenem, or amikacin) is also noteworthy. Although the clinical use of ceftazidime/avibactam has contributed to the development of KPC variants and the rapid progression of ceftazidime/avibactam resistance, combining ceftazidime/avibactam with another antimicrobial agent, for example, a carbapenem, may present an alternative therapeutic strategy for ceftazidime/avibactam-resistant KPC variant-producing CRKP strains. Most isolates were resistant to individual antibiotics used in combination. There were some isolates sensitive to both antibiotics used in some antimicrobial agent combinations. In these cases, these combinations exhibited indifferent effects on CRKP strains. But it should be noted that polymyxin-meropenem exhibited additive effects on 7 CRKP strains with KPC variants which were sensitive to polymyxin and meropenem, hinting that antimicrobial agent combinations can serve a crucial function in infections caused by CRKP strains with KPC variants and more effective combination therapy should be explore.

Metallo-β-lactamases constitute another common mechanisms for carbapenem resistance in CRKP strains (Zhang et al., 2017). The combination of ceftazidime/avibactam with aztreonam is a promising treatment option for metallo-β-lactamase-producing CRKP strains (Marshall et al., 2017). Previous studies have shown that this combination’s therapeutic effectiveness against bloodstream infections caused by metallo-β-lactamase-producing CRKP strains is superior to other active antibiotics, particularly regarding 30-day all-cause mortality (Falcone et al., 2021).

Our data showed that synergistic activity of the ceftazidime/avibactam plus aztreonam combination was observed in eight out of nine metallo-β-lactamase-producing CRKP strains (88.9%). These eight strains exhibited high resistance to aztreonam, with MIC values ranging from 32 to 256 μg/mL. The remaining NDM-5-producing CRKP strain showed indifferent activity with the ceftazidime/avibactam plus aztreonam combination, due to its susceptibility to aztreonam. This suggests that combining ceftazidime/avibactam is particularly effective in treating infections caused by metallo-β-lactamase-producing CRKP strains with high MICs for both ceftazidime/avibactam and aztreonam.

OXA-48-like enzymes are identified as the third most prevalent carbapenemases among CRKP isolates. Polymyxin-based combination antimicrobial regimens demonstrated comparatively high synergy rates against OXA-48-like-producing CRKP strains. These results are consistent with findings from Balkan’s study, which indicates that polymyxin-based dual combinations lead to significantly better treatment outcomes compared to non-polymyxin based combinations in patients with bloodstream infections caused by OXA-48-like carbapenemase-producing Enterobacteriaceae (Balkan et al., 2014). Furthermore, the combination of meropenem and cefepime combination exhibited more favorable effects on OXA-48-like-producing CRKP strains than on KPC-2-producing CRKP strains, potentially due to the difference in affinity for meropenem and cefepime between KPC-2 enzymes and OXA-48-like enzymes. This highlights the need for further in-depth studies on the mechanisms underlying these synergistic effects.

However, there are some limitations of this study. Firstly, the number of CRKP isolates analyzed were relatively small due to the difficulty in obtaining KPC variant-producing CRKP strains. More samples should be included in our future research endeavors. Additionally, it remains uncertain whether the antimicrobial agent combinations that demonstrate good synergistic or addictive effects in vitro will also have effective anti-infection impacts in vivo. The in vivo effects and mechanisms need to be further explored in future studies. Thirdly, the effectiveness of more complicated combinations (like polymyxin, aztreonam and meropenem) aren’t explored. Future research should address this aspect to provide new strategies for treatment of infectious diseases.

In summary, this study has investigated highly effective antimicrobial agent combinations against CRKP isolates producing various carbapenemases. The combinations of polymyxin-aztreonam, polymyxin-meropenem and polymyxin-levofloxacin demonstrated superior synergistic effects and could serve as effective therapeutic options for carbapenemase-producing CRKP strains. Additionally, ceftazidime/avibactam-based dual combinations show promise as reliable treatment regimens against KPC variant-producing CRKP strains. To ensure the viability of these combinations in clinical settings, further mechanistic studies and more extensive clinical trials are urgently needed.