Aztreonam-avibactam: an option against carbapenem-resistant Enterobacterales with emerging resistance

Dear Editor,

The Enterobacterales are an order of Gram-negative bacteria comprising a few major human pathogens such as Escherichia coli and Klebsiella pneumoniae. However, carbapenem-resistant Enterobacterales (CRE) has risen as an urgent threat for human health, leading to high mortality with very limited antimicrobial options. The main mechanism mediating resistance to β-lactams including carbapenems in the Enterobacterales is production of β-lactamases, which are two categories of enzymes capable of hydrolyzing β-lactams: serine β-lactamases and metallo-β-lactamases (MBLs). Avibactam (AVI) is a non-β-lactam β-lactamase inhibitor able to inhibit almost all serine β-lactamases but not MBLs. AVI in combination with ceftazidime (CAZ) has been approved for treating infections caused by CRE but CAZ-AVI has no activities against those producing MBLs. Currently, no MBL inhibitors have been approved for clinical use. Aztreonam (ATM), a monobactam, is stable to the hydrolysis of MBLs, and AVI can protect ATM from the inactivation by serine β-lactamases. The ATM-AVI combination may therefore be a viable choice against CRE producing serine β-lactamases, MBLs and/or both, which have been supported by large-scale in vitro susceptibility studies.¹

At present, ATM-AVI has not been approved for clinical use. Nevertheless, the combination of ATM and CAZ-AVI, both of which are available in clinical settings, can form de facto the ATM-AVI combination. An precision medicine approach can be detection of the type of carbapenem-hydrolyzing β-lactamases (serine β-lactamases and/or MBLs) followed by the selection of CAZ-AVI against CRE without producing MBLs or CAZ-AVI in combination with ATM for those producing MBLs. Such ATM-CAZ-AVI combination has been recommended by European Society of Clinical Microbiology and Infectious Diseases (ESCMID) and Infectious Diseases Society of America for treating severe infections caused by MBL-producing CRE.²,³ Currently available limited data suggest that ATM-CAZ-AVI is commonly associated with lower mortality rate, fewer clinical failure, and shorter length of stay compared with other active agents such as colistin-based combinations.^4–6 A systematic review⁴ including 91 patients with infections of MBL-producing CRE from 16 case reports, case series, and cohorts has indicated that 79% of them had clinical resolution within 30 days after receiving ATM-CAZ-AVI. A later retrospective study⁵ including 22 patients receiving ATM-CAZ-AVI against MBL-producing CRE showed a 18.2% 30-day mortality.

However, resistance to ATM-AVI in Enterobacterales has also emerged. Notably, both the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST, https://www.eucast.org) have not specified the breakpoints for defining resistance to ATM-AVI. Nevertheless, the CLSI resistance breakpoint for ATM (minimum inhibitor concentration [MIC] ≥16 mg/L) is commonly applied for ATM-AVI with AVI fixed at 4 mg/L in literature. Studies of mechanisms for resistance to ATM-AVI have been mainly targeted E. coli and K. pneumoniae. In E. coli, penicillin binding protein 3 (PBP3) is the major target of ATM and a four-amino-acid insertion (YRIK, YRIN, YRIP, or TIPY) in PBP3 could render a reduced susceptibility background against ATM.⁷ In such a background, a few β-lactamases able to efficiently hydrolyze ATM, such as AmpC cephalosporinase CMY-42,⁷,⁸ extended-spectrum β-lactamases CTX-M-15 and CTX-M-199, and carbapenemase KPC-21, could evade the protection from AVI and mediate resistance to ATM-AVI (Table 1). In particular, increased production of these β-lactamases in the presence of PBP3 with a four-amino-acid insertion could lead to elevated levels of resistance to ATM-AVI.⁷

Table 1.

Aztreonam-avibactam-resistant mechanisms in Escherichia coli and Klebsiella pneumonia.

Strain	Country of isolation	Speciesa	STb	MICc, ATM-AVI, mg/L	Resistance mechanismd
035123	China	EC	410	16	PBP3 ins (YRIK) + CMY-42
035148	China	EC	410	128	PBP3 ins (YRIK) + CMY42 overexpression
035166R	China	EC	410	512	PBP3 ins (YRIN) + KPC-21
Survcare058	Germany	EC	405	16	PBP3 ins (YRIK) + CMY-42
Survcare453	Germany	EC	410	16	PBP3 ins (YRIK) + CMY-42
Survcare253	Germany	EC	2851	16	PBP3 ins (YRIN) + CMY-42
ARC3799	India	EC	410	16	PBP3 ins (YRIN) + CMY-42
R-3058	Angola	EC	-	16	PBP3 ins (YRIN) + CMY-42
R-461	France	EC	-	16	PBP3 ins (YRIN) + CMY-42
1114251	Turkey	EC	410	16	PBP3 ins (YRIK) + CMY-42
1126350	Vietnam	EC	617	16	PBP3 ins (YRIN) + CMY-42
WCHEC96200	China	EC	405	16	PBP3 ins (YRIK) + CTX-M-15
PF_19–40f	UK	EC	-	16	BaeS (Q163L), likely with OmpC mutatione
PF_19–51e	UK	EC	-	16	BaeS (Y42H), likely with OmpC mutatione
PF_19–51n	UK	EC		32	BaeS (G241D), likely with OmpC mutatione
PF_19–52d	UK	EC	-	16	BaeR (G23E), likely with OmpC mutatione
PF_19–52j	UK	EC		16	BaeS (F159L), likely with OmpC mutatione
PF_19–26a	UK	EC	-	16	CMY-44 (Y150C)
91471	China	KP	11	128	KPC-2 overexpression + OmpK alterationf
96202	China	KP	11	128	KPC-2 overexpression + OmpK alterationf
98180	China	KP	11	64	KPC-2 overexpression + OmpK alterationf
98690	China	KP	11	32	KPC-2 overexpression + OmpK alterationf
108728	China	KP	11	>128	KPC-2 overexpression + OmpK alterationf
108738	China	KP	11	>128	KPC-2 overexpression + OmpK alterationf
108783	China	KP	11	32	KPC-2 overexpression + OmpK alterationf
109096	China	KP	11	>128	KPC-2 overexpression + OmpK alterationf
116216	China	KP	11	64	KPC-2 overexpression + OmpK alterationf
1116221	Thailand	KP	273	>16	acrA overexpression + DHA-1 + OmpK36 (Y43X)
Kp202_32A	China	KP	-	>256	CMY-16 (Y150S) + ∆OmpK35 + OmpK36 (ins)
Kp214-R150	China	KP	147	128	CMY-16 (Y150S) + OmpK35 (ISEcp1) + OmpK36 (ins)
Kp231-R150	China	KP	377	128	CMY-16 (Y150S) + OmpK35 (IS1)
Kp518-R150	China	KP	258	32	CMY-16 (Y150S) + ∆OmpK35
Kp202_64A	China	KP	-	>256	CMY-16 (N346H) + ∆OmpK35 + OmpK36 (ins)
Kp214-R346	China	KP	147	64	CMY-16 (N346H) + OmpK35 (ISEcp1) + OmpK36 (ins)
Kp231-R346	China	KP	377	32	CMY-16 (N346H) + OmpK35 (IS1)
Kp518-R346	China	KP	258	16	CMY-16 (N346H) + ∆OmpK35
KP21(ramR) ompK36 pKPC-3 V239G	UK	KP	-	16	ramR (frameshift) + ∆OmpK36 + KPC-3 (V239G)

^aEC,E. coli; KP,K. pneumoniae.

^b-, not determined.

^cMIC, in the presence of 4 mg/L AVI.

^dins, insertion of amino acids. IS1 and ISEcp1 are two insertion sequences interrupting OmpK35. ∆, truncation. The letters in parentheses represent amino acid insertion (YRIK or YRIN) or amino acid substitution with the location being indicated (e.g. S106P refers the substitution of S at location 106 by P). BaeS/BaeR, transcription activator of multidrug efflux pump-encoding genes mdtABCD and acrD; CMY, AmpC cephalosporinase; CTX-M, an extended-spectrum β-lactamase; DHA-1, AmpC cephalosporinase; KPC, Klebsiella pneumoniae carbapenemase; OmpK, outer membrane porin; PBP3, penicillin binding protein 3;ramR, repressor of ramA related to efflux pump and porins.

^eThe details of OmpK36 mutations are not available for these strains.

^fOmpK alteration refers to decreased expression of OmpK35 + truncated OmpK37 + absence of OmpK36.

Mechanisms without the involvement of PBP3 have also been identified. Unlike E. coli, alteration of penicillin-binding proteins has not been identified in ATM-AVI-resistant K. pneumoniae. In contrast, alteration and/or decreased production of outer membrane porins OmpK35 (OmpF), OmpK36 (OmpC), and/or OmpK37 could also provide the reduced susceptibility background against ATM in Enterobacterales, particularly in K. pneumonia.^8–10 Alteration of these porins may result from various mechanisms such as mutation, truncation, insertion/deletion of amino acids, or interruption by insertion sequences,^8–10 while their decreased production is largely due to mutations in regulator genes such as BaeS-BaeR two-component system and ramR, both of which also regulate efflux pumps. In the background of decreased production and/or alteration of porins, overexpression of carbapenemase gene bla_KPC-2 could confer resistance to ATM-AVI and the resistance level correlates to the expression of bla_KPC-2 in K. pneumonia.⁹ KPC-3 with a V239G (Ambler position) amino acid substitution alone does not confer resistance to ATM-AVI but is able to confer such resistance in the presence of truncated OmpK36 and frameshift mutation of ramR, which also could result in overexpression of the AcrAB-TolC efflux pump and decreased production of OmpK35 in K. pneumonia.¹⁰ Overexpression of the AcrAB-TolC efflux pump has also been found to confer resistance to ATM-AVI in combination with the production of DHA-1 (an AmpC cephalosporinase) and OmpK36 with mutation in K. pneumonia.⁸ However, whether overexpression of efflux pumps truly plays a role in mediating resistance to ATM-AVI remains to be determined owing to the co-existence of multiple other potential mechanisms in single strains. For CMY β-lactamases, there are eight AVI binding residues (S64, K67, E120, Y150, N152, K315, T316, and N346; one letter of amino acid and Ambler position). CMY with mutations occurring at these residues such as Y150C, Y150S and N346H could lead to ATM-AVI resistance alone in E. coli or in the presence of porin alteration in K. pneumoniae (Table 1).

In summary, ATM-AVI, which can be realized by combining ATM with CAZ-AVI at present, provides the full coverage for CRE producing MBLs and/or serine-β-lactamases. Currently-available limited clinical data suggest that ATM-AVI leads to favorable clinical outcomes and has been recommended as the first-line choice against MBL-producing CRE by well-recognized guidelines. Alarmingly, CRE resistant to ATM-AVI has also emerged. Resistance to ATM-AVI in Enterobacterales are multifactorial, mainly due to the combination of two or three of the following mechanisms: 1) enhanced hydrolysis to ATM and/or reduced binding to AVI due to overproduction or mutations of β-lactamases able to hydrolyze ATM; 2) decreased membrane permeability due to alterations, absence, or reduced production of porins such as OmpC/OmpF in E. coli or OmpK35/OmpK36 in K. pneumoniae; 3) target modification mainly due to modification of PBP3 in E. coli; and possibly 4) increased drug efflux. The clinical efficacy of ATM-AVI warrants further studies with large-scale patient sizes. Rigorous monitoring is also needed for ATM-AVI resistance in CRE.

Conflict of interest

None to declare.