Klebsiella pneumoniae carbapenemase variants: the new threat to global public health

SUMMARY

Klebsiella pneumoniae carbapenemase (KPC) variants, which refer to the substitution, insertion, or deletion of amino acid sequence compared to wild bla_KPC type, have reduced utility of ceftazidime-avibactam (CZA), a pioneer antimicrobial agent in treating carbapenem-resistant Enterobacterales infections. So far, more than 150 bla_KPC variants have been reported worldwide, and most of the new variants were discovered in the past 3 years, which calls for public alarm. The KPC variant protein enhances the affinity to ceftazidime and weakens the affinity to avibactam by changing the KPC structure, thereby mediating bacterial resistance to CZA. At present, there are still no guidelines or expert consensus to make recommendations for the diagnosis and treatment of infections caused by KPC variants. In addition, meropenem-vaborbactam, imipenem-relebactam, and other new β-lactam-β-lactamase inhibitor combinations have little discussion on KPC variants. This review aims to discuss the clinical characteristics, risk factors, epidemiological characteristics, antimicrobial susceptibility profiles, methods for detecting bla_KPC variants, treatment options, and future perspectives of bla_KPC variants worldwide to alert this new great public health threat.

INTRODUCTION

Antimicrobial resistance has become one of the most severe threats to global public health. Infections caused by carbapenem-resistant Enterobacterales (CRE) are associated with a particularly significant economic burden and mortality rate (1), which is often two to three times higher than carbapenem-susceptible Enterobacterales (2). Furthermore, CRE isolates are resistant to most of the antimicrobial agents available, limiting the choice of antimicrobial agents in clinical practice (2–4). In 2017, the World Health Organization published a priority list of antibiotic-resistant bacteria that listed CRE, carbapenem-resistant Pseudomonas aeruginosa (CRPA), and carbapenem-resistant Acinetobacter baumannii (CRAB) as the most urgent threats, suggesting that there is an urgent need to develop new antimicrobial agents to counteract the rapid acquisition of antimicrobial resistance (5).

The most important mechanism underlying carbapenem resistance in carbapenem-resistant Klebsiella pneumoniae (CRKP) and carbapenem-resistant Escherichia coli is the production of carbapenemases, including class A carbapenemases [mainly Klebsiella pneumoniae carbapenemase (KPC)], class B metallo-β-lactamases [mainly New Delhi metallo-β-lactamase (NDM)] , and some class D OXA-48-like carbapenemases (mainly OXA-181, OXA-232, and OXA-163) (6). Based on those, novel β-lactam-β-lactamase inhibitor combinations were developed continually to cope with infections caused by carbapenemase-producing Enterobacterales, including ceftazidime-avibactam (CZA), meropenem-vaborbactam, and imipenem-relebactam. CZA, launched in the USA in 2015 and in China in 2019, displayed potent in vitro activity against KPC-producing Enterobacterales and is a pioneer antimicrobial agent in treating infections caused by KPC-producing Enterobacterales (7–9). Unlike traditional enzyme inhibitors such as tazobactam and sulbactam, avibactam does not contain a β-lactam structure. Avibactam acts by covalent acylation of its β-lactamase targets in a reversible process in which the structure of avibactam is restored by deacylation (without hydrolysis), and intact avibactam is released to provide long-lasting enzyme inhibition effect (10). Most importantly, avibactam has a broader spectrum, inhibiting class A carbapenemases (particularly KPC-2), extended-spectrum β-lactamases (ESBLs), class C cephalosporinases, and some class D carbapenemases. Therefore, since its introduction into clinical use, CZA has been considered one of the most effective antimicrobial agents for the treatment of infections caused by KPC-producing strains, especially K. pneumoniae (9).

Nevertheless, the widespread clinical use of CZA has forced CRE to mutate to adapt to the increasing antibiotic pressure. The bla_KPC variant derived from bla_KPC-2 or bla_KPC-3 mutation has been reported (11–13). The bla_KPC variant usually refers to the substitution, insertion, or deletion of one or more amino acids compared to wild bla_KPC type (such as bla_KPC-2 and bla_KPC-3), and those leading to modifications in the amino acid sequence with the carbapenemase active site, which are of greatest concern. So far, more than 145 bla_KPC variants have been reported worldwide, and most of the new variants were discovered in the past 3 years. A crucial phenotypic feature of KPC variants is their resistance to CZA compared to the wild-type (WT) gene product that is notably susceptible. This has led to new challenges in appropriate therapeutic selection (11, 14). In addition to posing a challenge for antimicrobial therapy, bla_KPC variants may challenge the performance of some classical carbapenemase detection methods used by clinical laboratories (14).

This review aims to discuss the clinical characteristics, risk factors, epidemiological characteristics, antimicrobial susceptibility profiles, methods for detecting bla_KPC variants, treatment options, and future perspectives of bla_KPC variants worldwide to alert this new significant public health threat.

EFFICACY OF CEFTAZIDIME-AVIBACTAM IN THE TREATMENT INFECTIONS CAUSED BY KPC-PRODUCING ENTEROBACTERALES

A multicenter, prospective, observational study was conducted between January 2018 and March 2019 to evaluate the outcomes and predictors of mortality in patients with KPC- or OXA-48-K. pneumoniae infections treated with CZA, with a focus on KPC-K. pneumoniae bloodstream infections (BSIs) (15). The study included 147 patients, 140 isolates were KPC-positive and 7 isolates were OXA-48-positive. Of these patients, 68 (46.3%) received CZA and 79 (53.7%) received CZA in combination with at least one other active agent, such as colistin, aminoglycosides, colistin plus tigecycline, tigecycline, trimethoprim/sulfamethoxazole, fosfomycin, etc. The 14- and 28-day mortality rates were 9% and 20%, respectively. The clinical characteristics of 71 patients with KPC-K. pneumoniae BSIs treated with a regimen containing CZA (cases) and of 71 patients, matched by propensity score, whose KPC-K. pneumoniae BSIs were treated with regimens not containing ceftazidime/avibactam (controls) were collected in this study. The results showed that 28-day mortality in the patients with KPC-K. pneumoniae BSIs treated with CZA was significantly lower than that observed in the matched patients whose KPC-K. pneumoniae BSIs were treated with regimens without CZA (18.3% vs 40.8%; P = 0.005). A retrospective, observational, multicenter study of 577 adults with BSIs or non-bacteremic infections primarily involving the urinary tract, lower respiratory tract, and intra-abdominal structures was conducted to evaluate the outcomes and mortality in patients with KPC-K. pneumoniae infections treated with CZA monotherapy and CZA combination regimens (16). The results showed that the 30-day all-cause mortality rate was 25% (146/577). There was no significant difference in mortality between patients treated with CZA alone and those treated with combination regimens (26.1% vs 25.0%, P = 0.79). Another retrospective multicenter study of 138 Italian patients with KPC-K. pneumoniae infections (KPC-K. pneumoniae bacteremic, KPC-K. pneumoniae non-bacteremic), all of whom received CZA as salvage therapy, was conducted to document the clinical characteristics and outcomes of these cases and to specifically investigate the outcomes and predictors of mortality in patients with KPC-K. pneumoniae bacteremia (17). All KPC-K. pneumoniae isolates were susceptible to CZA in vitro. The overall 30-day mortality rate was 34.1% (47/138), and 8.7% (12/138) of patients experienced recurrent KPC-K. pneumoniae infection after discontinuation of CZA treatment. In addition, the 30-day mortality rate of 109 patients with KPC-K. pneumoniae bacteremia receiving CZA was significantly lower than that of the control group (36.5% vs 55.7%, P = 0.005). This suggests that CZA-based therapy is an effective treatment option for infections caused by KPC-producing Gram-negative bacteria.

In the context of increasing carbapenem resistance in Enterobacterales, the introduction of CZA, with its broad spectrum of enzyme inhibition and low side effects, has provided a major boost to the clinical treatment of CRE infections (7, 9). CZA is the preferred treatment option for KPC-producing infections outside of the urinary tract (18). However, as CZA becomes more widely used in clinical practice, KPC-producing strains develop resistance to CZA through mutation under conditions of antimicrobial pressure and the presence of various other factors, leading to treatment failure (19, 20).

EPIDEMIOLOGICAL CHARACTERISTICS OF KPC-TYPE CARBAPENEMASE

First identified in 1996 in the northeastern United States, KPCs are the most common and widely distributed carbapenemases (21). KPCs can hydrolyze most β-lactams, including carbapenems, cephalosporins, monobactams, and cephamycins. They have been identified in many Gram-negative bacteria, including Enterobacterales and nonfermenting bacteria (e.g., P. aeruginosa and A. baumannii). K. pneumoniae is the most predominant KPC-producing species (22, 23). The strains producing KPC-2 or KPC-3 showed diverse susceptibility to imipenem and meropenem but were usually resistant to ertapenem. With the widespread use of carbapenems, KPC-producing bacteria have spread internationally. The dissemination of bla_KPC involves clonal spread, horizontal transfer, and plasmid-mediated spread. The prevalent bla_KPC variants vary with geographic location, e.g., a predominance of bla_KPC-2 in China but a predominance of bla_KPC-3 and bla_KPC-2 in the Americas and Europe (24–28). The international spread of KPC-producing K. pneumoniae is primarily associated with a single multilocus sequence type, sequence type 258 (ST258), and its variants (29). K. pneumoniae ST258 is the culprit in over 77% of KPC-producing K. pneumoniae infections in the USA and 90% in Israel (25, 26). The profile of sequence types (STs) was different between regions. ST512 is the most prevalent ST in Italy (28), whereas ST11 and ST437 predominate in China and Brazil, respectively (30, 31). Transmission of bla_KPC genes is mainly mediated by horizontal transfer of plasmids or mobility of small genetic elements (mainly Tn4401 transposon) (32). The bla_KPC in K. pneumoniae had been reported on various plasmid types, such as IncF, IncI2, IncX, IncA/C, IncR, and ColE1, but the predominant plasmid type is IncF with FII K replicon (29, 33, 34).

Sporadic variants of bla_KPC were identified in 2006–2018 (≤7 reported per year) (35). However, since 2019, reports of bla_KPC variants have increased dramatically. The number reported between 2019 and 2020 exceeds the sum of reported bla_KPC variants in the previous 17 years (Fig. 1 and 2). The emergence of bla_KPC-2 or bla_KPC-3 mutants, such as bla_KPC-33, bla_KPC-31, etc., as the most prevalent cause of bacterial resistance to CZA during therapy has been attributed to the mechanism that altered the Ω-loop hydrogen bonding structure of KPC (Fig. 3) (36, 37). Compared to the wild type, KPC variant protein presented reduced catalytic ability to carbapenems, meanwhile responding poorly to avibactam inhibition (38–40). The crystallographic analysis of the D179N KPC-2 variant and the D179Y variant revealed the depth mechanism of CZA resistance conferred by D179N/Y variants of KPC-2 (40). The D179N KPC-2 structure showed that the change of the carboxyl to an amide moiety at position 179 disrupted the salt bridge with R164 present in KPC-2. Additional interactions were disrupted in the Ω-loop, causing a decrease in the melting temperature. Shifts originating from N179 were also transmitted toward the active site, including ∼1-Å shifts of the deacylation water and interacting residue N170 (40). The structure of the D179Y KPC-2 (KPC-33) revealed more drastic changes, as this variant exhibited disorder of the Ω-loop, with other flanking regions also being disordered (40). On the contrary, the crystal structure of D179N KPC-2 in complex with vaborbactam revealed wild-type KPC-2-like vaborbactam-active site interactions, with D179N change in KPC-2 that does not perturb the binding mode of vaborbactam significantly (40).

Fig 1.

Distribution of Klebsiella pneumoniae carbapenemase variants based on KPC-2 and KPC-3 mutations. The data are calculated based on the number of Klebsiella pneumoniae carbapenemase variants uploaded to the National Center for Biotechnology Information each year in each region (https://www.ncbi.nlm.nih.gov/).

Fig 2.

KPC variants evolutionary tree analysis chart (A) and the number of new variants reported in PubMed over the years (B) (up to March 2023).

Fig 3.

Ω-Loop destabilization as a mechanism of resistance to ceftazidime-avibactam. (A) Structure of KPC-2 (Protein Data Bank accession number 5UL8), Ω-Loop was shown in yellow; (B) hydrogen bonding network involving Ω-loop Arg and Asp residues in the WT KPC-2 (40); (C) Hydrogen bonding network involving Ω-loop Arg and Asp residues in D179N KPC-2 (40).

By March 2023, 145 bla_KPC variants are registered in the National Center for Biotechnology Information (NCBI) database, all derived from mutations of bla_KPC-2 or bla_KPC-3. Overall, bla_KPC-4 (n = 26), bla_KPC-33 (n = 9), bla_KPC-12 (n = 8), bla_KPC-6 (n = 5), bla_KPC-71 (n = 4), bla_KPC-10 (n = 3), bla_KPC-76 (n = 3), bla_KPC-44 (n = 3), bla_KPC-25 (n = 2), bla_KPC-36 (n = 2), bla_KPC-5 (n = 2), and bla_KPC-90 (n = 2) were mutants from bla_KPC-2. These KPC-2 variants were mainly identified from the USA (n = 35), China (n = 20), and Italy (n = 3). In contrast, bla_KPC-31 (n = 8), bla_KPC-66 (n = 4), bla_KPC-67 (n = 3), bla_KPC-18 (n = 2), bla_KPC-29 (n = 2), bla_KPC-40 (n = 2), bla_KPC-49 (n = 2), bla_KPC-61 (n = 2), and bla_KPC-70 (n = 2) were mutants from bla_KPC-3. These bla_KPC-3 variants were mainly reported from Italy (n = 15), the USA (n = 8), and France (n = 1). We aggregated the number of bla_KPC variants uploaded to NCBI each year by region, and the global distribution of specific variants is shown in Fig. 1. The USA and China ranked first and second with 63 and 59 detected cases, respectively. It is worth noting that China showed a sudden increase in bla_KPC variants in 2020 after ceftazidime-avibactam was approved for clinical use in 2019. The threat of bla_KPC variants in China may be more worthy of vigilance. The bla_KPC variants gene have been identified in various Gram-negative bacilli, including K. pneumoniae (73.8%), Enterobacter hormaechei (7.1%), E. coli (6%), and Enterobacter cloacae (5.5%), among which K. pneumoniae is predominant (Tables 1 and 2).

TABLE 1.

Distribution of KPC variants based on bla_KPC-2

KPC-2-like enzyme	Country and year of first publication	Geographical distribution	Bacteria	Divergence from KPC-2	Plasmids	Clonal dissemination	Genetic environment	Accession number
KPC-4	USA, 2013	USA	Enterobacter cancerogenus, Enterobacter cloacae, Serratia marcescens, K. pneumoniae	P104R/V240G	IncL/M, IncN, IncHI2	K. pneumoniae: ST258, ST834, ST964, ST307; Enterobacter cloacae: ST171	Tn4401 variants	JQ837276
KPC-5	USA, 2013	USA	Pseudomonas aeruginosa, K. pneumoniae	Pro103→Arg	IncX	K. pneumoniae: ST429	Tn5563 and IS6100A, Tn4401 variants	NG_049259
KPC-6	USA, 2012	USA	K. pneumoniae	Val240Gly			Tn4401 variants	EU555534
KPC-10	Puerto Rico, 2010	Puerto Rico	Acinetobacter calcoaceticus- baumannii	H272Y				GQ140348
KPC-12	China, 2021	China	K. pneumoniae	L169M	IncFII, IncR		Downstream of ISKpn28 and upstream of ISKpn6	CP036400
KPC-14	France, 2019	France, USA, Italy	K. pneumoniae	D242-GT-243 deletion	IncN	K. pneumoniae: ST16, ST1685	Tn4401b	CP045022
KPC-15	China, 2014	China	K. pneumoniae	119 Leucine and 146 lysine			ISKpn6-like transposon	KC433553
KPC-16	China, 2016	China	K. pneumoniae	P202S and F207L		K. pneumoniae: ST11		NG_049249
KPC-17	China, 2016	China	K. pneumoniae	F207L		K. pneumoniae: ST11		KC465200
KPC-21	Portugal, 2018	Portugal	E. coli	Trp105Arg		E. coli: ST131	Upstream a partial ISKpn6 and downstream a truncated Tn3 transposon	MH133192
KPC-24	Chile, 2016	Chile	K. pneumoniae	R6P		K. pneumoniae: ST1161	Tn4401a	KR052099
KPC-33	Italy, 2020	Italy, China, Greece	K. pneumoniae	D179Y	IncFII-R	K. pneumoniae: ST11, ST1685, ST39		MT550691
KPC-35	USA, 2019	USA	K. pneumoniae	L169P	IncF	K. pneumoniae: ST258		MH404098
KPC-35	China, 2019	China	K. pneumoniae	L169P		K. pneumoniae: ST11		CP050279.1
KPC-36	Italy, 2020	Italy	K. pneumoniae	D163E		K. pneumoniae: ST1519	Tn4401a	MK976712
KPC-44	Greece, 2021	Greece, Finland	K. pneumoniae	276ins YTRAPNKDDKHSEAV	IncFIB	K. pneumoniae: ST39	Tn4401a	SAMN14734455
KPC-51	China, 2020	China	K. pneumoniae	D179N, Y241H, H274N				MN725731
KPC-52	China, 2020	China	K. pneumoniae	D179Y, valine insertion after 262 position				MN725732
KPC-55	South Korea, 2020	South Korea	K. pneumoniae	T794A	IncX3	K. pneumoniae: ST307	Tn4401a	MT028409
KPC-57	Greece, 2021	Greece	K. pneumoniae	D179V	IncFIB	K. pneumoniae: ST39	Tn4401a	MT358626
KPC-71	China, 2022	China	K. pneumoniae	Ser182dup	pKpQIL	K. pneumoniae: ST11	Tn4401a	OK315339
KPC-74	China, 2021	China	K. pneumoniae	G239_V240del	IncFII-R	K. pneumoniae: ST11	TnAs1-ISKpn6-like-blaKPC– 74-ISKpn27-TniA-IS26	NG_070742
KPC-82	USA, 2021	USA	Citrobacter koseri	Ala267-Ser275	IncN		Tn4401 variants	MW485086
KPC-90	China, 2021	China	Pseudomonas aeruginosa	180ins Tyr-Thr		Pseudomonas aeruginosa: ST463	IS26-ISKpn6-blaKPC-90-IS26	MZ570431
KPC-93	China, 2022	China	K. pneumoniae	267ins Pro-Asn-Asn-Arg-Ala	IncFII-R	K. pneumoniae: ST11	Tn1721-ΔISKpn6- blaKPC-93-ISKpn8-IS26	MZ569034
KPC-112	China, 2022	China	K. pneumoniae	del166Leu/167Asn and 242Gly/243Thr	IncFII-R	K. pneumoniae: ST15	Tn1721-ΔISKpn6- blaKPC-112-ISKpn8-IS26	OM177660
KPC-113	China, 2023	China	Pseudomonas aeruginosa	266ins Gly	Type I	P. aeruginosa: ST3903	IS6100-ISKpn27-blaKPC-113- ΔISKpn6-Tn1403	OM317762
KPC-123	China, 2022	China	Citrobacter koseri	ins179_TY and ins270_DDKHSEA			ISKpn27-blaKPC-ISKpn6	ON012820

TABLE 2.

Distribution of KPC variants based on bla_KPC-3

KPC-3-like enzyme	Country and year of first publication	Geographical distribution	Bacteria	Divergence from KPC-3	Plasmids	Clonal dissemination	Genetic environment	Accession number
KPC-23	Greece, 2019	Greece	K. pneumoniae	V240A	IncFIIk-FIB	K. pneumoniae: ST258	Tn4401a	AWU66461
KPC-28	France, 2019	France	E. coli	Δ242-GT-243 and H274Y				KY282958
KPC-29	Italy, 2021	Italy	K. pneumoniae	D272insKDD	pKpQIL	K. pneumoniae: ST512		WP_096807439
KPC-31	Italy, 2019	Italy	K. pneumoniae	D179Y	FIA(HI1)-R	K. pneumoniae: ST512, ST307, ST101, ST2502	Tn4401 variants	MAPH01000113
KPC-39	Italy, 2020	Italy, France	K. pneumoniae	A172T	FIA(HI1)-R	K. pneumoniae: ST101
KPC-40	USA, 2019	USA	Enterobacter hormaechei	L167_E168dup	IncN	Enterobacter hormaechei: ST407		WP_115470049
KPC-41	Switzerland, 2019	Switzerland,	K. pneumoniae	269-Pro-Asn-Lys-270	IncFII	K. pneumoniae: ST395		MK497255
KPC-48	Spain, 2020	Spain	K. pneumoniae	L169P-A172T				MN422013
KPC-49	Spain, 2021	Spain, Italy	E. coli, K. pneumoniae	Arg-163-Ser	IncF	E. coli: ST131, K. pneumoniae: ST512	Tn4401a	MN619655
KPC-50	Switzerland, 2020	Switzerland	K. pneumoniae	276-Glu-Ala-Val-277	IncFIB	K. pneumoniae: ST258		MN654342
KPC-94	Spain, 2022	Spain	K. pneumoniae	LN169-170H		K. pneumoniae: ST512
KPC-95	Spain, 2022	Spain	K. pneumoniae	D179Y and A172T		K. pneumoniae: ST512
KPC-115	Argentina, 2022	Argentina	K. pneumoniae	del168Leu/169Asn and Ser170Pro		K. pneumoniae: ST11	Tn4401a	OM714909
KPC-121	Italy, 2022	Italy	K. pneumoniae	179ins Ser		K. pneumoniae: ST512		SAMN27596901
KPC-125	Italy, 2022	Italy	K. pneumoniae	Asp179Ala		K. pneumoniae: ST512		ON012820

The most frequently reported bla_KPC-2 variant, bla_KPC-33, was derived from bla_KPC-2 with a single amino acid substitution. It was first reported in Italy in October 2020 and subsequently reported in China and Greece (41, 42). In these studies, bla_KPC-33 was harbored by K. pneumoniae, which was predominantly isolated from rectal swabs and sputum samples. Although the ST distribution of KPC-33-producing K. pneumoniae varies in different regions, ST11 in China, ST1685 in Italy, and ST39 in Greece, these strains share the same feature that all bla_KPC-33 are carried by Tn4401-like transposons (30, 37, 42). The most frequently reported bla_KPC-3 variant, bla_KPC-31, derived from bla_KPC-3 with a single amino acid substitution (D179Y), was first reported in Italy in August 2019 and subsequently mainly reported in Italy, France, and Spain (12, 36, 43–45). The K. pneumoniae strain producing bla_KPC-31 is primarily detected in blood specimens, and the corresponding STs included ST512, ST307, ST101, and ST2502. Like other bla_KPC variants, all bla_KPC-31 are located on the genetic element of Tn4401-like transposons (12, 36, 43–45).

THE CLINICAL CHARACTERISTICS AND RISK FACTORS OF bla_KPC VARIANTS

In 2017, Shields et al. (46) first reported the evolution of bla_KPC in three patients with infection due to KPC-producing K. pneumoniae, resulting in therapy failure. bla_KPC-3 was mutated to bla_KPC-31 and bla_KPC-32 during CZA treatment for 10 to 19 days. By searching the PubMed database, up to August 2022, 18 cases of infections caused by KPC variant-producing Enterobacterales have been reported internationally. These infections were reported in 11 males, 5 females, and 2 patients of unknown sex. K. pneumoniae was the pathogen in 16 cases, E. coli and C. koseri in one case each (Table 3) (11, 44–55). Limited data showed that bla_KPC-31 (44.4%, 8/18) and ST 258 (22.2%, 4/18) were the dominant genotypes and ST type of bla_KPC variants. All patients were treated with CZA (for 9–41 days) before KPC variant-producing Enterobacterales strains were identified, suggesting that CZA treatment may be an independent risk factor for inducing mutations of bla_KPC. In the 18 patients infected with a KPC variant-producing CRE strain, the treatment failure rate was up to 38.9% (7/18), and 50% (8/16) of the patients were older than 50. Four patients continued to receive CZA in combination with other antimicrobial agents (including meropenem, tigecycline, amikacin, or imipenem) for anti-infective treatment after CZA-resistant bla_KPC variant strains were identified. Among them, three patients were not successfully cured after receiving the above anti-infection regimens, while one was cured after receiving CZA in combination with imipenem. Eleven patients discontinued CZA therapy and were treated with other antimicrobial agents, to which the bacterial isolates were susceptible, including tigecycline, polymyxin B, colistin, meropenem, and amikacin. Among them, seven patients’ infection symptoms were effectively controlled. Three patients were successfully treated with meropenem-vaborbactam, based on the bacteria being in vitro susceptible to meropenem-vaborbactam. However, more real-world research data are still needed to confirm the effectiveness of meropenem-vaborbactam in treating infection caused by the bla_KPC variant.

TABLE 3.

Summary of KPC variant cases

No.	Sex/age, years	Infection	Organism	Wild-type	KPC variant	ST	Therapy before the KPC variant was detected	Therapy	Outcome	Pathogen clearance	Reference
1	Male/59	Pneumonia	K. pneumoniae	KPC-3	KPC-31	ST2502	CZA, 10 d	Meropenem 2 g tid 3 h	Cure	No	(45)
2	Male/68	Bloodstream infection	K. pneumoniae	KPC-3	KPC-31	ST512	CZA,14 d	Meropenem-vaborbactam 1/1 g q8	Cure	Yes	(44)
3	Male/40	Bloodstream infection, pneumonia	K. pneumoniae	KPC-3	KPC-31	ST307	Meropenem, tigecycline, polymyxin B, CZA	Meropenem + polymyxin B	Failure	No	(51)
4	Female/50+	Subphrenic abscess	K. pneumoniae	KPC-3	KPC-8, KPC-31	ST258	CZA, 19 d	Piperacillin/tazobactam, gentamicin	NA	No	(46)
5	Female/70+	Pneumonia	K. pneumoniae	KPC-3	KPC-31	ST258	CZA, 15 d	Meropenem + polymyxin B	Cure	Yes	(46)
6	Female/NA	Bloodstream infection, pneumonia	K. pneumoniae	KPC-3	KPC-31	ST1519	CZA, 14 d	Meropenem (increased dose, extended infusion time) + gentamicin	Failure	No	(50)
7	Female/60	Pneumonia, wound infection	K. pneumoniae	KPC-3	KPC-31	ST101	CZA, 14 d	CZA + tigecycline	Failure (death)	No	(48)
8	Male/47	Intra-abdominal infection	K.pneumoniae	KPC-3	KPC-31/KPC-39/KPC-47/KPC-48	NA	CZA, 12 d	Imipenem, tigecycline, gentamicin	Cure	No	(53)
9	Male/67	Bloodstream infection, pneumonia	K. pneumoniae	KPC-3	KPC-3 (A177E, D179Y)	ST258	CZA, 15 d	Meropenem 1 g q12, 12 d	Cure	Yes	(52)
10	Female/40+	Pneumonia, urinary tract infection	K. pneumoniae	KPC-3	KPC-32	ST258	CZA, 14 d	Meropenem	Failure	No	(46)
11	Male/24	Bloodstream infection, pneumonia	K. pneumoniae	KPC-2	KPC-33	NA	CZA, 33 d	Polymyxin B + gentamicin + tigecycline, meropenem- vaborbactam 25 d	Cure	Yes	(49)
12	Male/42	Pneumonia	K. pneumoniae	KPC-2	KPC-33	ST11	CZA, 16 d	Imipenem + amoxicillin/clavulanate + polymyxin B	Failure	No	(11)
13	NA	Bloodstream infection	K. pneumoniae	KPC-2	KPC-44	ST39	CZA + tigecycline, 14 d	Colistin + trimethoprim/sulfamethoxazole	Cure	Yes	(47)
14	Female/40+	Bloodstream infection	E. coli	KPC-3	KPC-49	ST131	CZA, 10 d	Amikacin 500 mg/24 h i.v.	Cure	Yes	(54)
15	NA	Bloodstream infection	Citrobacter koseri	NA	KPC-82	NA	CZA, 28 d	Meropenem-vaborbactam	Cure	Yes	(56)
16	Male/22	Pneumonia	K. pneumoniae	KPC-2	KPC-33	ST11	CZA, 41 d	Imipenem + CZA	Cure	Yes	(57)
17	Female/52	Pneumonia	K. pneumoniae	KPC-2	KPC-79, KPC-76	ST11	CZA, 9 d	CZA + amikacin, meropenem + tigecycline + fosfomycin	Failure	No	(55)
18	Male/50	Pneumonia	K. pneumoniae	KPC-2	KPC-35, KPC-78, KPC-33	ST859	CZA, 9 d	CZA + tigecycline + fosfomycin + meropenem	Failure	No	(55)

A retrospective, observational research was conducted with all patients admitted to the intensive care unit (ICU) dedicated to coronavirus disease patients at the City of Health & Sciences in Turin, between May 2021 and January 2022, with the primary endpoint to study strains with resistance to CZA. This study enrolled 17 patients with colonization or invasive infection caused by CZA-resistant K. pneumoniae which were susceptible to meropenem. It is worth noting that 76.5% of patients did not receive therapy with CZA. Cluster analysis showed that 16 KPC-33-producing K. pneumoniae isolates belonged to a single clone, indicating that there was a risk of clonal transmission of this novel bla_KPC variant (58).

KPC-2 or KPC-3 carbapenemase-producing K. pneumoniae are susceptible to mutations during treatment leading to treatment failure, and the occurrence of such mutations is not easily recognized clinically and may be missed, putting infected patients at high risk. Available studies suggest that the use of ceftazidime-avibactam is an important factor in the development of mutations in KPCase-producing K. pneumoniae. As there are no current guidelines for the treatment of infections caused by KPCase-producing mutants of K. pneumoniae, it is recommended that clinical monitoring for changes in antimicrobial susceptibility to antimicrobial agents, particularly carbapenems and ceftazidime-avibactam, be performed at regular intervals (e.g., 3–5 days) to monitor for the occurrence of mutations.

GENETIC SEQUENCES SURROUNDING bla_KPC CARBAPENEMASE GENES

Transposable elements play an essential role in bacteria’s genetic variation and evolution. In most countries and regions, such as Europe (33), the USA (59), and Brazil (31, 60), the bla_KPC is mainly located on transposable elements like Tn4401, Tn3-Tn4401 chimera CTB, and IS26. Tn4401 and Tn3-Tn4401 chimera CTB, belonging to the Tn3 family, can mobilize bla_KPC-2 at a high transposition frequency. Tn4401, a 10 kb transposon, has been reported as the genetic structure mediating original bla_KPC acquisition, with the gene order of Tn4401-tnpR, Tn4401-tnpA, ISKpn7, bla_KPC, and ISKpn6 (61). Owing to the diversity in the intervening sequence between the ISKpn7 and bla_KPC genes, a total of eight unique Tn4401 isoforms (a to h) have been characterized, with Tn4401a and Tn4401b being the most widespread (62). In Asia, bla_KPC-2 is predominantly located on different variants of Tn1721 and IS26 (63–65). However, the structural sequences surrounding the bla_KPC variants genes are infrequently described. Scattered studies have shown that some KPC variants, similar to bla_KPC-2 and bla_KPC-3, are located within the Tn4401 transposon (56). The core structure of ISKpn6-bla_KPC- ISKpn28 was identified in bla_KPC-12 and bla_KPC-74 (Tables 1 and 2). Therefore, we collected 52 bla_KPC variant gene sequences from NCBI databases to analyze the genetic environment surrounding the bla_KPC gene. The result showed the core structure of the tnpR-tnpA-ISKpn7-bla_KPC-ISKpn6 was identified in the strains carrying bla_KPC-31, bla_KPC-32, bla_KPC-33, bla_KPC-14, bla_KPC-56, bla_KPC-29, bla_KPC-68, bla_KPC-27, bla_KPC-49, or bla_KPC-115. The core structure of the ISKpn27-bla_KPC-ISKpn6 was identified in the strains producing bla_KPC-71, bla_KPC-79, bla_KPC-93, bla_KPC-123, bla_KPC-51, or bla_KPC-52 (Fig. 4).

Fig 4.

Genetic sequences surrounding bla_KPC carbapenemase genes.

SUSCEPTIBILITY PROFILE OF THE STRAINS PRODUCING KPC VARIANTS

The bla_KPC variants can cause a seesaw effect. On the one hand, under the pressure of CZA, the bla_KPC sequence mutates, including bla_KPC-8 (66), bla_KPC-30, bla_KPC-31 (36), bla_KPC-32, bla_KPC-40 (67), bla_KPC-41 (68), bla_KPC-50 (69), and bla_KPC-57 (42), resulting in changes in the protein structure and weakening its ability to bind to avibactam, thus making the bacteria resistant to ceftazidime-avibactam [minimum inhibitory concentration (MIC) range 16–>128 mg/L]. On the other hand, the bacteria regained their susceptibility to carbapenems (especially imipenem, Table 4). In recent years, with the increasing public attention to bla_KPC variants, studies have found that some bla_KPC variants play avibactam-resistant ESBL profile, including bla_KPC-14 (70), bla_KPC-28 (71), bla_KPC-33 (36), bla_KPC-46, bla_KPC-51 (72), bla_KPC-52 (72), bla_KPC-53 (73), and bla_KPC-66 (36). The ESBL phenotype-positive KPCs can be inhibited by the classic enzyme inhibitor clavulanic acid; ESBL phenotype confirmatory tests that routinely use clavulanic acid as an inhibitor often report positive results, misleading the clinicians to consider the strains as ESBL producers rather than carbapenemase producers. KPC variants producing strains tended to be inhibited by meropenem-vaborbactam. Therefore, meropenem-vaborbactam is considered a salvage therapy after failure of CZA treatment (49). It is important to note that CRE strains carrying bla_KPC variants with decreased susceptibility to CZA can also show cross-resistance to the siderophore cephalosporin cefiderocol (74). This may be related to the fact that cefiderocol and ceftazidime are very similar in structure.

TABLE 4.

Antimicrobial susceptibility profiles and carbapenemase assay results of KPC variant-producing strains^a

Organism	Variants	MIC (mg/L)								Carbapenemase assay						Reference
		ETP	IPM	MEM	CZA	MEV	IMR	COL	TGC	mCIM	Rapid Carba NP	NG-Test Carba 5	RESIST-5 O.O.K.N.V	GeneXpert	RAPIDEC Carba NP
K. pneumoniae	KPC-12	256(R)	16(R)	64(R)	4(S)	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND	(75)
NA	KPC-14	ND	ND	ND	ND	ND	ND	ND	ND	NEG	NEG	POS	ND	POS	NEG	(71)
K. pneumoniae	KPC-14	ND	≤1(S)	≤0.125(S)	64(R)	ND	ND	ND	ND	NEG	NEG	POS	POS	ND	ND	(12)
K. pneumoniae	KPC-23	ND	512(R)	512(R)	16(R)	4(S)	ND	1(S)	4(I)	ND	ND	ND	ND	ND	ND	(76)
E. coli	KPC-28	ND	ND	ND	ND	ND	ND	ND	ND	NEG	NEG	POS	ND	POS	NEG	(71)
K. pneumoniae	KPC-31	>1(R)	<0.25(S)	2(I)	>8(R)	0.125(S)	ND	0.5(S)	0.5(S)	ND	ND	ND	ND	ND	ND	(44)
K. pneumoniae	KPC-31	ND	≤1(S)	4(R)	>256(R)	ND	ND	ND	ND	NEG	NEG	NEG	NEG	ND	ND	(12)
K. pneumoniae	KPC-31	ND	0.25(S)	4(R)	>64(R)	1(S)	ND	0.5(S,PB)	2(S)	ND	ND	ND	ND	ND	ND	(41)
K. pneumoniae	KPC-33	32(R)	0.5(S)	4(R)	>64(R)	2(S)	ND	ND	ND	NEG	ND	ND	ND	POS	ND	(14)
K. pneumoniae	KPC-35	16(R)	1(S)	4(R)	64(R)	2(S)	ND	ND	ND	NEG	ND	POS	ND	POS	ND	(14)
K. pneumoniae	KPC-36	>32(R)	>256(R)	>256(R)	16(R)	2(S)	ND	ND	2(S)	ND	ND	POS	ND	POS	ND	(77)
K. pneumoniae	KPC-39	ND	>16(R)	>16(R)	>16(R)	16(R)	8(R)	0.5(S)	1(S)	ND	ND	ND	ND	ND	ND	(53)
K. pneumoniae	KPC-41	4(R)	4(R)	1(S)	>128(R)	ND	ND	ND	ND	ND	POS	ND	ND	ND	ND	(68)
K. pneumoniae	KPC-47	ND	>16(R)	>16(R)	>16(R)	16(R)	16(R)	1(S)	1(S)	ND	ND	ND	ND	ND	ND	(53)
K. pneumoniae	KPC-48	ND	1(S)	1(S)	>16(R)	1(S)	1(S)	1(S)	1(S)	ND	ND	ND	ND	ND	ND	(53)
E. coli	KPC-49	ND	2(I)	0.5(S)	16(R)	ND	ND	≤2(S)	≤0.5(S)	ND	ND	ND	ND	ND	ND	(54)
K. pneumoniae	KPC-50	1(I)	16(R)	2(I)	>256(R)	<0.125(S)	2(S)	ND	ND	ND	POS	ND	ND	ND	ND	(69)
K. pneumoniae	KPC-51	ND	1(S)	4(R)	2048(R)	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND	(72)
K. pneumoniae	KPC-52	ND	2(I)	2(I)	256(R)	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND	(72)
K. pneumoniae	KPC-53	ND	2(I)	4(R)	64(R)	2(S)	ND	>8(R)	1(S)	ND	ND	ND	POS	ND	ND	(73)
K. pneumoniae	KPC-55	ND	1(S)	2(I)	0.125(S)	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND	(78)
K. pneumoniae	KPC-70	18 mm (R)	28 mm (S)	21 mm (I)	>256(R)	ND	ND	ND	22 mm (S)	ND	ND	ND	ND	ND	ND	(79)
K. pneumoniae	KPC-71	16(R)	0.5(S)	2(I)	>64(R)	0.5(S)	ND	ND	ND	NEG	ND	NEG	ND	POS	ND	(14)
K. pneumoniae	KPC-74	16(R)	0.5(S)	1(S)	128(R)	ND	ND	ND	8(R)	ND	ND	NEG	ND	ND	ND	(80)
K. pneumoniae	KPC-76	32(R)	2(I)	4(R)	>64(R)	2(S)	ND	ND	ND	NEG	ND	NEG	ND	POS	ND	(14)
K. pneumoniae	KPC-78	32(R)	0.5(S)	2(I)	>64(R)	1(S)	ND	ND	ND	NEG	ND	POS	ND	POS	ND	(14)
K. pneumoniae	KPC-79	32(R)	4(R)	8(R)	64(R)	4(S)	ND	ND	ND	NEG	ND	POS	ND	POS	ND	(14)
Citrobacter koseri	KPC-82	ND	4(R)	2(I)	128(R)	0.125(S)	ND	ND	ND	ND	ND	ND	ND	POS	ND	(56)
K. pneumoniae	KPC-94	>2(R)	≤0.5(S)	2(I)	>16(R)	ND	ND	0.5(S)	1(S)	ND	ND	ND	ND	POS	ND	(81)
K. pneumoniae	KPC-95	>2(R)	1(S)	1(S)	>16(R)	ND	ND	0.5(S)	0.5(S)	ND	ND	ND	ND	POS	ND	(81)
K. pneumoniae	KPC-112	ND	0.5(S)	1(S)	>128(R)	ND	ND	ND	2(I)	ND	ND	ND	ND	ND	ND	(82)
K. pneumoniae	KPC-115	0.5(S)	≤0.25(S)	≤0.5(R)	24(R)	ND	ND	ND	ND	ND	ND	POS	ND	ND	ND	(83)
K. pneumoniae	KPC-121	≥256(R)	8(R)	16(R)	≥256(R)	32(R)	8(R)	0.125(S)	ND	ND	ND	ND	ND	ND	ND	(84)
Citrobacter koseri	KPC-123	≤0.5(S)	≤0.5(S)	≤0.5(S)	64(R)	ND	ND	1(S)	≤0.25(S)	ND	ND	ND	ND	ND	ND	(85)

^a MIC, minimum inhibitory concentration; KPN, K. pneumoniae; ECO, E. coli; CKO, Citrobacter koseri; ETP, ertapenem; IPM, imipenem; MEM, meropenem; CZA, ceftazidime-avibactam; MEV, meropenem-vaborbactam; IMR, imipenem-relebactam; COL, colistin; PB, polymyxin B; TGC, tigecycline; mCIM, modified carbapenem inactivation method; POS, positive; NEG, negative; ND, not detected.

KPC VARIANTS POSE NEW CHALLENGES FOR LABORATORY TESTING

Rapid detection of bla_KPC is essential in treating CRE infections. Standard methods for detecting carbapenemases include both phenotypic and genotypic assays. Phenotypic assays include the Carba NP assay, modified carbapenem inactivation method (mCIM), EDTA-modified carbapenem inactivation method (eCIM), 3-aminophenylboronic acid (APB)/EDTA method, and time-of-flight mass spectrometry (86–88). Genotypic detection methods based on nucleic acid detection techniques, include GeneXpert Carba R assay (Cepheid, Sunnyvale, CA, USA), Verigene Gram-negative blood culture test (Nanosphere, Northbrook, IL, USA), and FilmArray system (bioMérieux, Marcy l’Étoile, France) (89–92). Phenotypic assays are mainly based on the ability of carbapenemases to hydrolyze carbapenems. The sensitivity and specificity of Carba NP assay, mCIM, eCIM, and APB/EDTA method are higher than 90% in detecting carbapenemases [including KPC, NDM, Verona metallo-β-lactamase (VIM), São Paulo metallo-β-lactamase (SPM)] (93–97). Moreover, phenotypic assays have been highly favored due to their low cost. However, the emergence of bla_KPC variant genes poses significant challenges to such phenotypic assays in detecting bla_KPC variant genes among CZA-resistant CRE (14). Since KPC-variant producing strains are often susceptible or intermediate to imipenem or meropenem, carbapenemase phenotypic assays are prone to reporting false-negative results (81). NG-Test Carba 5 and RESIST-5 O.O.K.N.V are two of the most commonly used enzyme-linked immunochromatography-based test strips. Ding et al. (14) showed that NG-Test Carba 5 effectively detected bla_KPC-35 (n = 3), bla_KPC-78 (n = 1), and bla_KPC-79 (n = 1), but false-negative results were observed for bla_KPC-33 (n = 5), bla_KPC-71 (n = 1), and bla_KPC-76 (n = 8). A similar result was obtained by Bianco et al. in which NG-Test Carba 5 and RESIST-5 O.O.K.N.V could detect bla_KPC-14 (n = 2), but could not detect bla_KPC-31 (n = 4) and bla_KPC-33 (n = 2) (12). The false-negative results of enzyme immunochromatography in detecting bla_KPC variants may be related to the mutation site, which changes the target site of carbapenemase binding to the antibody and thus fails to bind to the antibody, resulting in false-negative results.

KPC variants may challenge the performance of some classical carbapenemase detection methods used by clinical laboratories. PCR assay may be suitable to detect bla_KPC variant. The GeneXpert Carba R assay is a qualitative, in vitro real-time PCR assay designed to detect five arbapenemases gene families, including bla_IMP, bla_{KPC-2 or 3}, bla_NDM, bla_OXA-48, and bla_VIM, with more than 96% sensitivity and specificity (90, 92, 98). Ding et al. (14) showed that GeneXpert Carba R has potential advantage in detecting bla_KPC variant due to its detection principle, which is not affected by gene locus variation and can effectively detect bla_KPC-33 (n = 5), bla_KPC-35 (n = 3), bla_KPC-71 (n = 1), bla_KPC-76 (n = 8), bla_KPC-78 (n = 1), and bla_KPC-79 (n = 1). Several studies showed that the presence of the novel bla_KPC-46, bla_KPC-66, bla_KPC-92, bla_KPC-94, and bla_KPC-95 were confirmed by the X-pert Carba R assay (81, 99). A similar study compared three carbapenemase detection methods, including GeneXpert Carba R, NG-Test Carba 5, and colloidal gold immunoassay test, to evaluate the performance of these methods in detecting KPC variants protein. The result showed that GeneXpert Carba R can confirm the presence of all 13 types of KPC variant protein, including KPC-2, KPC-3, KPC-25, KPC-33, KPC-35, KPC-51, KPC-52, KPC-71, KPC-76, KPC-77, KPC-78, KPC-93, and KPC-123 (100). There are other methods for detecting carbapenemases based on PCR amplification, such as Verigene Gram-negative blood culture test (Nanosphere, Northbrook, IL, USA), FilmArray system (bioMérieux, Marcy l’Étoile, France), etc., but there are no published data on the performance of these methods in detecting KPC variants (91). Different methods for detecting KPC variants may get different results. The laboratories need to pay attention to this special phenomenon. When the conventional carbapenemase detection method shows negative results, but the bacteria show resistance to CZA, further molecular testing is required to determine whether bacteria produce carbapenemase.

The emergence of KPC variants disrupts conventional laboratory thinking about carbapenemase detection and changes the practice of inferring bacterial susceptibility to CZA from carbapenemase detection results (most automated susceptibility testing systems do not yet include CZA). The reporting of false-negative carbapenemase results is likely to mislead clinicians in their anti-infective treatment. Based on the above information, we believe that CZA susceptibility testing should be performed concurrently with carbapenemase testing, and KPC variants can be identified early by combining the antimicrobial susceptibility phenotype and carbapenemase results to provide more information early to start precision treatment of infections caused by KPC variants.

CURRENTLY AVAILABLE TREATMENT OPTIONS

Carbapenems

Since several KPC-variant producing strains regained susceptibility to imipenem or meropenem while remaining resistant to CZA, this suggests the potential value of carbapenems in the treatment of infections caused by KPC-variant strains. Sporadic cases of treatment success have been reported for meropenem alone (increased dose, extended infusion time) or in combination with other antimicrobial agents in managing infections caused by KPC variant-producing K. pneumoniae (45, 52). However, meropenem or other antimicrobial combinations have also shown a high failure rate in treating such resistant strains in some cases when the bla_KPC variant reverts to the original bla_KPC-2 during treatment, allowing the bacteria to regain carbapenem resistance (11, 51, 57). Carbapenem therapy was associated with a 50% all-cause mortality rate in patients infected with KPC variant producer (usually related to clinical failure), which is much higher than that observed in patients with ESBL-producing K. pneumoniae infections and similar to that observed in KPC-2/3-producing K. pneumoniae infections not treated with CZA (53). The limited data currently available suggest that most bla_KPC variants occur during CZA treatment. As previously described, CZA inhibits KPC-2 or KPC-3, while meropenem or imipenem provides potent activity against most KPC variants. Since the bacteria may revert to classical bla_KPC phenotypes such as bla_KPC-2 or bla_KPC-3 producer in response to environmental changes when treated with carbapenems alone to which the bacterial isolates were susceptible, is it possible that an antimicrobial regimen of carbapenems combined with CZA could effectively cover both the KPC-2/3 and KPC variants producing strains, thus preventing the chance of clonal transformation of bacteria? Should we try this approach to meet the clinical needs? However, such combinations in treating KPC variant-producing bacterial infections are rarely reported (57). The efficacy and safety must also be validated by further data from in vitro or in vivo studies.

New β-lactam-β-lactamase inhibitor combinations

Meropenem-vaborbactam

Vaborbactam is a non-β-lactam serine β-lactamase inhibitor based on a cyclic boronic acid pharmacophore (101). Meropenem-vaborbactam inhibits the activity of class A serine enzymes (including ESBLs and KPC) and AmpCs, but not OXA-48 carbapenemases and metallo-β-lactamases (102). Meropenem-vaborbactam was approved by the U.S. Food and Drug Administration in 2017 for treating complicated urinary tract infections (cUTIs) in adults and by the European Medicines Agency in 2018 for treating cUTIs, complicated intra-abdominal infections, or hospital-acquired pneumonia and bacteremia occurring in association (or suspected association) with any of these infections (103). In the European Union, meropenem-vaborbactam is also indicated for treating infections caused by aerobic Gram-negative bacteria in adults with limited options. In a study by Lapuebla et al. evaluating the activity of meropenem-vaborbactam against KPC-producing Enterobacterales, the combination inhibited 98.5% (131/133) of isolates at 1/8 mg/L (104). A global multicenter epidemiological study in 2014 included 10,426 Enterobacterales strains, ≤2/8 mg/L of meropenem-vaborbactam inhibited 99.3% of the strains tested (105). Subsequent studies on 11,559 Enterobacterales strains collected in 2015 found that meropenem-vaborbactam inhibited 99.5% of KPC-producing Enterobacterales at a concentration of 4/8 mg/L (106). Studies have also shown potent in vitro activity of meropenem-vaborbactam against the K. pneumoniae strains producing KPC variants, including KPC-14, KPC-28, KPC-31, KPC-33, KPC-35, KPC-39, KPC-50, KPC-71, KPC-76, KPC-78, or KPC-79 (14, 107, 108). A retrospective study collected 12 KPC variant-producing K. pneumoniae isolates from ICU patients between November 2020 to January 2021, including KPC-62 (n = 11) and KPC-31 (n = 1) (109). The result showed that all strains that carried KPC variant were resistant to CZA (MIC ≥64 mg/L), but susceptible to meropenem-vaborbactam (MIC range, 0.25/8 mg/L ~ 2/8 mg/L) (109). Case reports have shown that meropenem-vaborbactam can successfully cure bacteremia caused by KPC-31 producers, suggesting that meropenem-vaborbactam may be a new option for treating infections caused by KPC variant-producing Enterobacterales strains (49). Similarly, meropenem-vaborbactam has successfully treated infections caused by KPC-82-producing Citrobacter koseri (56).

Imipenem-relebactam

Relebactam is a new β-lactamase inhibitor with a diazabicyclooctane core, similar to avibactam (110). Imipenem-relebactam inhibits class A serine enzymes (including ESBLs and KPC) and AmpCs, but not OXA-48 carbapenemase, metallo-β-lactamase, or class A carbapenemase GES-20 (111). The Study for Monitoring Antimicrobial Resistance Trends reported that relebactam restored susceptibility to imipenem in 80.5%, 100%, and 74.1% of imipenem-nonsusceptible Pseudomonas aeruginosa, Enterobacteriaceae, and K. pneumoniae, in 2015 (112). A study by Papp-Wallace et al. showed that all Enterobacterales strains were highly susceptible to imipenem-relebactam (MIC ≤2 mg/L) (113). A study by Carpenter et al. also showed similar results, with imipenem-relebactam being the most active combination against CRE (MIC_50/90 ≤0.25/0.5 mg/L) (114). In vitro studies have also shown excellent activity of imipenem-relebactam against the K. pneumoniae strains producing KPC variants, including KPC-31, KPC-33, KPC-35, and KPC-62 (109, 115). Therefore, it may be speculated that imipenem-relebactam has good in vitro antimicrobial activity against KPC variants. At the same time, imipenem-relebactam has good antimicrobial activity against KPC-2- or KPC-3-producing Enterobacterales, so in theory, using imipenem-relebactam in treating KPC variant infection can effectively prevent the emergence of KPC-2/3-producing strains.

Aztreonam-avibactam

The infections caused by metallo-β-lactamase-producing strains remain a major challenge in clinical treatment because avibactam, vaborbactam, and relebactam can not inhibit the activity of metallo-β-lactamases. However, the aztreonam-avibactam combination is highly anticipated because metallo-β-lactamases do not hydrolytically destroy the antimicrobial activity of aztreonam, and so aztreonam-avibactam can inhibit the Enterobacterales strains producing either class A serinease, metallo-β-lactamase, or class D OXA-48 carbapenemase (116). Interestingly, some KPC variant-producing strains regained susceptibility to carbapenems while simultaneously regaining susceptibility to aztreonam. KPC-31, KPC-33, KPC-49, and KPC-94 producers showed high susceptibility to aztreonam (MIC range: ≤0.5–4 mg/L), but KPC-95-, KPC-82-, KPC-55-, and KPC-14-producing strains remained resistant to aztreonam (MIC >16 mg/L) (54, 70, 81). Our results showed (Li Ding, et al. data unpublished) that the K. pneumoniae strains producing KPC variant (including KPC-33, KPC-35, KPC-71, KPC-76, KPC-78, KPC-79, and KPC-112) were highly susceptible to aztreonam-avibactam (MIC ranges, 2/4–4/4 mg/L).

Other antimicrobial agents

Tigecycline

Tigecycline belongs to a new class of glycylcycline antibiotics. It has been touted as one of the last lines of defense in treating complex infections caused by multi-drug resistant Gram-negative and Gram-positive bacteria. Binding to bacterial 30S ribosomes prevents the entry of transfer RNA. It prevents amino acids from integration into peptide chains, ultimately blocking bacterial protein synthesis and limiting bacterial growth. It has great in vitro antimicrobial activity against class A, B, C, and D β-lactamase-producing Enterobacterales (117). Scattered studies have shown that tigecycline has high in vitro antimicrobial activity against KPC variant-producing Enterobacterales (55, 82). But an excess mortality risk was demonstrated in comparative clinical trials (118). While it is generally not recommended for treating bacteremia because of its low steady-state concentrations in serum following the current dosing recommendation, tigecycline is primarily used in combination regimens when treating carbapenem-resistant Gram-negative infections (119). However, the emergence of tigecycline-resistant strains has recently been reported (120). Overexpression of resistance-nodulation-division efflux pumps such as AcrAB is an essential molecular mechanism underlying tigecycline resistance (121). Additionally, tet(X) gene variants are newly emerging mechanisms of tigecycline resistance (122, 123).

Polymyxin

Polymyxins are an “old” class of lipopeptide antibiotics approved in the late 1950s (124). Polymyxins have regained public attention due to their excellent activity against CRE, CRPA, and CRAB. The two clinically available polymyxins, colistin and polymyxin B, demonstrate comparable spectra of antibacterial activity, mechanism of action, and resistance profile because of their similar structures (125). CHINET report in 2018 showed that polymyxin B had excellent in vitro activity against 272 clinical isolates of CRKP (93.8% susceptible) (126). However, nephrotoxicity and heteroresistance are two major limitations of polymyxins (125, 127).

Modification of lipid A portion of lipopolysaccharide (LPS) is the main mechanism of Enterobacterales resistant to polymyxin, which can be caused by chromosomal mutations in genes of the two-component system involved in LPS modification, namely PhoPQ, PmrAB, and CrrAB, and MgrB (128). Additionally, plasmid-mediated polymyxin resistance gene mcr-1 can also medicate strains resistant to polymyxin (129). Several studies showed that KPC variants did not affect the susceptibility of polymyxin, suggesting that polymyxin still has potential advantages in treating KPC variant infections (41, 55).

FUTURE OUTLOOK

To address the challenges posed by the global spread of bla_KPC variants, multiple studies need to be conducted to curb the infections caused by these bacteria. Firstly, countries need to establish a collaborative surveillance network for the bla_KPC to survey the spread of KPC variant producers in real-time and carry out proactive hospital infection control measures to curb the spread. Secondly, more effective antimicrobial agents need to be developed continuously to deal with infections and be used rationally to avoid the emergence of the resistant strain. Countries should open fast-track approval channels for new drugs to be used to save patients as soon as possible. Thirdly, clinical microbiology laboratories should strengthen the routine detection and inform bla_KPC variant-positive isolates. In vitro diagnostic companies need to develop methods that can accurately and timely detect new bla_KPC variants. The manufacturers of automated antimicrobial susceptibility testing systems should add carbapenemase testing and report the result of KPC carbapenemases in advance based on clinical need. Rapid whole-genome sequencing should be applied to predict the resistance profile mediated by KPC variants. In the context of limited treatment options currently available, there is an urgent need to develop treatment guidelines for infections caused by KPC variant producers. Finally, ARC (avibactam-resistant carbapenemase) is classified as a class A enzyme. The designation of KPC variants is currently confusing. Some KPC variants are designated as ESBLs and some KPC variants are named as carbapenemases. More KPC variants are being reported, and more are expected in the future. Although the sequence difference between KPC variants and KPC wild type is not significant, the corresponding antimicrobial susceptibility profile and detection techniques are very different. The traditional KPC inhibitor APB cannot inhibit the activity of ARC. Such ARC enzymes should be named separately to attract attention.

Biographies

Li Ding received her Master's degree in Clinical Laboratory Diagnostics from Huashan Hospital, Fudan University, Shanghai, China in 2020. She is currently working in the Department of Clinical Microbiology at the Institute of Antibiotics, Huashan Hospital, Fudan University, mainly involved in bacterial resistance monitoring and in vitro pharmacodynamic evaluation of new antimicrobial agents. She is very interested in the mechanism of carbapenem-resistant Enterobacteriaceae and has published several articles in international peer-reviewed journals.

Siquan Shen received his Master's degree in Clinical Laboratory Diagnostics from Huashan Hospital, Fudan University, Shanghai, China in 2023. He is currently working in the Department of Clinical Microbiology, Institute of Antibiotics, Huashan Hospital. His main research interests include various aspects of carbapenemase-mediated resistance in enterobacterial species. He is also involved in projects on bacterial resistance monitoring research and in vitro pharmacodynamics of new antibiotics. Many related articles have been published in internationally renowned journals.

Jing Chen received his Master's degree from Southwest University in Chongqing, China in 2010. In the same year, he joined the BGI Microbial Platform, demonstrating a long-term commitment to the development of microbial NGS data. Over the years, he has been actively involved in numerous projects under the State High-Tech Development Plan (863 Programme) and the National Natural Science Foundation of China. Since 2019, he has been leading the development of bioinformatics-related algorithms and software at Hangzhou Matridx Biotechnology Co., Ltd.

Zhen Tian graduated from Fujian Agriculture and Forestry University, China with a Master's degree in Ecology in 2021. After graduation, he joined Hangzhou Matridx Biotechnology Co., Ltd. as a bioinformatics analyst, where he was responsible for the development of R-related processes. He has a strong interest in bioinformatics and next-generation sequencing (NGS), with a commitment to using bioinformatics tools to identify sources of infection.

Qingyu Shi received her Master's degree in Clinical Laboratory Diagnostics from Huashan Hospital of Fudan University in 2021, and went on to study for her PhD. She mainly researches antibiotic resistance mechanisms of enterobacteria, especially those mediated by carbapenemase. She is also actively involved in antimicrobial surveillance and in vitro pharmacodynamics of new antibiotics. Several related articles have been published in internationally renowned journals.

Renru Han works in clinical microbiology at the Institute of Antibiotics, Huashan Hospital, Fudan University. She works on bacterial resistance monitoring research and in vitro pharmacodynamics of new antibiotics. She has been focusing on the resistance mechanisms of carbapenem-resistant Enterobacterales for several years. Some related articles have been published in Journal of Antimicrobial Chemotherapy, Microbiology Spectrum, Frontiers in Microbiology, etc.

Yan Guo works in clinical microbiology at the Institute of Antibiotics, Huashan Hospital, Fudan University. She has focused on bacterial resistance surveillance research for nearly twenty years. She was a visiting scholar in Yohei Doi's lab at the University of Pittsburgh from 2016 to 2017. Another part of her work is in vitro pharmacodynamics of new antibiotics. Some related articles have been published in JAC, AAC, EJCMID, etc. She is skilled in antimicrobial susceptibility testing and is interested in research on bacterial resistance mechanisms, especially in VRE.

Fupin Hu is a clinical microbiologist at Huashan Hospital, Fudan University. He received his PhD in infectious diseases in 2010 and is now deputy director of the Institute of Antibiotics. Dr Hu has long been involved in research, mainly focusing on bacterial resistance surveillance and new mechanisms of drug resistance. He is one of the leaders of the China Antimicrobial Surveillance Network (CHINET, www.chinets.com). His main research interests include methodological standardisation of antimicrobial susceptibility testing, development of breakpoint for antimicrobial susceptibility testing of new antibacterial agents, evaluation of in vitro activities of new antimicrobial agents, mechanisms of transmission of bacterial resistance to carbapenems (especially mutation of KPC enzyme gene) and risk factors for infection.