ARGONAUT-IV: susceptibility of carbapenemase-producing Klebsiella pneumoniae to the oral bicyclic boronate β-lactamase inhibitor ledaborbactam combined with ceftibuten

ABSTRACT

Ledaborbactam (formerly VNRX-5236), a bicyclic boronate β-lactamase inhibitor with activity against class A, C, and D β-lactamases, is under development as an orally bioavailable etzadroxil prodrug (VNRX-7145) in combination with ceftibuten for the treatment of urinary tract infections. At ceftibuten breakpoints of ≤1 mg/L (EUCAST) and ≤8 mg/L (CLSI), 92.5% and 99.0%, respectively, of 200 carbapenem-resistant Klebsiella pneumoniae isolates, predominantly K. pneumoniae carbapenemase producing, were susceptible to ceftibuten-ledaborbactam (ledaborbactam tested at a fixed concentration of 4 mg/L) compared to 4.5% and 30.5%, respectively, to ceftibuten alone.

INTRODUCTION

Treatment of urinary tract infections (UTIs) due to multidrug-resistant Gram-negative pathogens has presented many clinical challenges, particularly when oral treatment is desired. Current Infectious Diseases Society of America Guidance on the treatment of uncomplicated cystitis caused by carbapenem-resistant Enterobacterales (CRE) includes oral ciprofloxacin, levofloxacin, trimethoprim-sulfamethoxazole or nitrofurantoin, or a single parenteral dose of an aminoglycoside, if isolates are susceptible to these agents (1). However, many isolates of CRE are multidrug-resistant, and there is a need for alternative oral agents. Only parenteral agents, ceftazidime-avibactam, meropenem-vaborbactam, imipenem-relebactam, and cefiderocol are recommended for treatment of pyelonephritis and complicated UTIs caused by CRE.

A discovery program combining medicinal chemistry, biochemical testing, microbiological profiling, and evaluation of oral bioavailability led to the development of ledaborbactam etzadroxil, which is the etzadroxil ester prodrug of ledaborbactam (2–4). The ester is cleaved after absorption to ledaborbactam, the active β-lactamase inhibitor. Ledaborbactam etzadroxil is under development in combination with ceftibuten, an orally bioavailable third-generation cephalosporin, for the treatment of complicated UTIs caused by Enterobacterales, including multidrug-resistant strains that produce serine β-lactamases (Ambler classes A, C, and D). Ceftibuten was FDA-approved in 1995 for the treatment of acute bacterial exacerbations of chronic bronchitis, acute bacterial otitis media, pharyngitis, and tonsillitis. At the time of its introduction into clinical use, ceftibuten had an in vitro spectrum of activity that included >90% of common species of Enterobacterales associated with UTIs (5, 6). A review of the activity of ceftibuten published in 1995 documented that the MIC₅₀ of ceftibuten against both Escherichia coli and Klebsiella pneumoniae was ≤0.25 mg/L and the MIC₉₀ was ≤0.5 mg/L (5). A more recent study of a 2018–2020 global culture collection of 3,889 clinical isolates of Enterobacterales showed a higher rate of resistance to ceftibuten of 33.1%, with modal MICs of ceftibuten susceptible isolates of 0.25–0.5 mg/L (7). Ceftibuten has been studied for the treatment of UTIs, both alone (8, 9) and in combination with amoxicillin-clavulanate to cover UTIs caused by extended-spectrum β-lactamase (ESBL)-producing species (10).

In vitro studies have demonstrated that ledaborbactam restored the activity of ceftibuten against Enterobacterales expressing Ambler class A extended-spectrum β-lactamases, class A carbapenemases, class C cephalosporinases, and class D oxacillinases (4). In vivo studies showed that ceftibuten-ledaborbactam dosed subcutaneously and ceftibuten-ledaborbactam etzadroxil dosed orally demonstrated similar activity, with median effective dose values of 13.5 and 12.9 mg/kg, respectively (11). In a murine ascending UTI model with three strains of E. coli expressing ESBLs (CTX-M-15) or a serine carbapenemase (KPC-2), ceftibuten-ledaborbactam in a 1:1 ratio resulted in increased efficacy, with bacterial titers that were ≥2.0 log₁₀ colony forming units (CFU)/mL lower in kidneys, ≥3.2 log₁₀ CFU/mL lower in bladders, and ≥4.0 log₁₀ CFU/mL lower in urine compared to treatment with ceftibuten alone (12).

This study was undertaken to evaluate the in vitro activity of ceftibuten-ledaborbactam against a carbapenem-resistant collection of Klebsiella pneumoniae with characterized carbapenem resistance mechanisms representative of those found in many regions of the USA (13–15). Study isolates included 200 clinical isolates collected 2012–2016 in the Great Lakes Region as part of the Antibacterial Resistance Leadership Group (ARLG) Consortium on Resistance against Carbapenems in Klebsiella (CRACKLE-I) Study (16). All isolates had been characterized as carbapenem-resistant, and all but one contained a known carbapenemase as determined by whole genome sequencing (NCBI BioProjects PRJNA339843 and PRJNA433394). Among these 200 isolates, 192 contained a KPC, six contained an OXA-48/OXA-48 variant, and one contained a variant of OXA-48 as well as a class B New Delhi metallo-β-lactamase (NDM) (17). MICs of ceftibuten-ledaborbactam and comparators, including ceftazidime-avibactam, cefepime-taniborbactam, and meropenem-vaborbactam, were determined by broth microdilution using Sensititre custom frozen panels (Thermo Fisher, Cleveland, OH), with ledaborbactam at a fixed concentration of 4 mg/L (18). MICs of the comparator agents were reported in a previous publication (17). MICs were interpreted using CLSI M100-S33 standards (18) with the following exceptions. (i) Tigecycline MICs were interpreted using FDA breakpoints for Enterobacterales (19). (ii) Ceftibuten MICs were interpreted using EUCAST UTI breakpoints (susceptible, ≤1 mg/L; resistant, >1 mg/L) (20) and CLSI UTI breakpoints (susceptible, ≤8 mg/L; intermediate, 16 mg/L; resistant, >1 mg/L). (iii) EUCAST UTI and CLSI breakpoints for ceftibuten were provisionally applied to ceftibuten-ledaborbactam. MIC interpretations for oral agents, ceftibuten and ceftibuten-ledaborbactam, are intended to be applied only to isolates causing UTIs, whereas interpretations for parenterally administered comparators apply to isolates causing systemic and urinary infections.

Of the 200 carbapenem-resistant K. pneumoniae isolates tested, 192 contained a class A K. pneumoniae carbapenemase (KPC), six contained a carbapenem-hydrolyzing oxacillinase (OXA-48) or an OXA-48-like variant, and one contained an OXA-48-like variant as well as a class B New Delhi metallo-β-lactamase (NDM). One isolate with CTX-M-15, OXA-1, TEM-1, and SHV-28 β-lactamases had no known carbapenemase. MIC_50/90 values and percentage of isolates susceptible to ceftibuten and ceftibuten-ledaborbactam are shown in Table 1, and a histogram of ceftibuten-ledaborbactam MICs by carbapenemase type is shown in Fig. 1. As expected, based on all but one of the isolates containing carbapenemases, susceptibility to ceftibuten was low (4.5% at EUCAST and 30.5% at CLSI breakpoints). Susceptibility to ceftibuten-ledaborbactam was high, with 92.5% susceptible at the EUCAST breakpoint and 99.0% susceptible at the CLSI breakpoint. At the EUCAST breakpoint, 10 of the 15 isolates resistant to ceftibuten-ledaborbactam had MICs of 2 mg/L, one doubling dilution above the ceftibuten susceptible breakpoint (Table 2 ). As previously reported, 98.0% of isolates were susceptible to ceftazidime-avibactam, 95.5% to meropenem-vaborbactam, 99.5% to cefepime-taniborbactam, 60.0% to amikacin, 77.0% to colistin, and 88.5% to tigecycline (17). Only seven isolates (3.5%) were susceptible to ciprofloxacin, which was tested in a previous study (16).

TABLE 1.

MIC_50/90 values and percent susceptibility of K. pneumoniae isolates (n = 200)^a

Agent (susceptible breakpoint, mg/L)	MIC range (mg/L)	MIC50 (mg/L)	MIC90 (mg/L)	Percent susceptible
Amikacin (≤16)	≤0.5 to >32	16	32	60.0
Colistin (≤2)b	0.25 to >4	0.5	>4	77.0
Ceftazidime (≤4)	0.5 to >16	>16	>16	1.0
Ceftazidime-avibactam (≤8)	≤0.06 to >8	1	2	98.0
Meropenem (≤1)	0.5 to >4	>4	>4	2.5
Meropenem-vaborbactam (≤4)	≤0.06 to 16	≤0.03	1	95.5
Tigecycline (≤2)	0.5 to >4	1	4	88.5
Ceftibuten (≤1/≤8)c	0.25 to >16	16	>16	4.5/30.5
Ceftibuten-ledaborbactam (≤1/≤8)d	≤0.12 to >16	≤0.12	1	92.5/99.0

^a MIC values of β-lactam/β-lactamase inhibitor combinations are shown as MICs of the β-lactam component in the presence of fixed concentrations of inhibitors: avibactam, 4 mg/L; vaborbactam 8 mg/L; ledaborbactam, 4 mg/L. Data on agents other than ceftibuten and ceftibuten-ledaborbactam have previously been published (17).

^b CLSI interpretation is intermediate with no susceptible category (18).

^c EUCAST/CLSI UTI breakpoints.

^d Using EUCAST/CLSI UTI breakpoints for ceftibuten.

Fig 1.

Histogram of ceftibuten-ledaborbactam MICs of K. pneumoniae by carbapenemase type. Upper panel shows MICs of isolates with KPC-2 and KPC-3. Lower panel shows MICs of isolates with other carbapenem resistance mechanisms.

TABLE 2.

Distribution of MICs of β-lactam/β-lactamase inhibitor combinations against K. pneumoniae isolates (n = 200)^a

Agent (susceptible breakpoint, mg/L)	Number of isolates with MIC value (mg/L) for agent shown
	≤0.06	0.12	0.25	0.5	1	2	4	8	16	32
Ceftazidime (≤8)	-b	-	-	2	-	-	-	1	2	195*c
Ceftazidime-avibactam (≤8)	7	6	12	55	60	45	10	1	4*	-
Meropenem (≤1)	-	-	-	2	3	9	21	165*	-	-
Meropenem-vaborbactam (≤4)	135	5	11	18	18	4	-	9*	-	-
Ceftibuten (≤1/≤8)d	-	-	3	1	5	6	17	39	60	69*
Ceftibuten-ledaborbactam (≤1/≤8)e	-	114	25	18	28	10	1	2	-	2*

^a MIC values of β-lactam/β-lactamase inhibitor combinations are shown as MICs of the β-lactam component in the presence of fixed concentrations of inhibitors: avibactam, 4 mg/L; vaborbactam 8 mg/L; ledaborbactam, 4 mg/L. Data on agents other than ceftibuten and ceftibuten-ledaborbactam have previously been published (17).

^b -, no isolates with MIC indicated or concentration not tested.

^c *, MIC value > next lowest MIC value.

^d EUCAST/CLSI UTI breakpoints.

^e Using EUCAST/CLSI UTI breakpoints for ceftibuten.

MICs of the β-lactam and β-lactam/β-lactamase inhibitor combinations tested against 14 isolates that were non-susceptible to ceftazidime-avibactam or meropenem-vaborbactam and/or with ceftibuten-ledaborbactam MICs > 2 mg/L, and all OXA-48/OXA-48-like producing isolates are shown in Table 3 with their β-lactam resistance mechanisms. Two of the meropenem-vaborbactam non-susceptible isolates, both expressing OXA-48, were susceptible to ceftazidime and ceftibuten alone, with MICs of ceftazidime-avibactam and ceftibuten-ledaborbactam within one doubling dilution of these agents alone. Six of the seven OXA-48/OXA-48-like producing isolates were non-susceptible to meropenem-vaborbactam, with three KPC-producing isolates also resistant to this agent. Four isolates were resistant to ceftazidime-avibactam (one OXA-48-like and three KPC-producing). Eight of these 14 isolates were susceptible to ceftibuten-ledaborbactam. Thirteen of these 14 isolates had mutations in ompK35 (n = 13), ompK36 (n = 8), and/or ftsI (encoding PBP3) genes (n = 1). The ompK35 mutations included insertion sequences (IS), deletions, and premature stop codons. The ompK36 mutations included two insertions known to reduce permeability of β-lactam agents (21). The ftsI mutation encoded a variant PBP3 (F383Y).

TABLE 3.

MICs of the β-lactam and β-lactam/β-lactamase inhibitor combinations tested against 14 isolates that were non-susceptible to ceftazidime-avibactam or meropenem-vaborbactam and/or with ceftibuten-ledaborbactam MICs > 2 mg/L, and all OXA-48/OXA-48-like producing isolates with their β-lactam resistance mechanisms^a

Isolate no.	Resistance mechanisms				MIC (mg/L) by agent
	β-Lactamases	Porin OmpK35	Porin OmpK36	PBP3b	Ceftazidime	Ceftazidime- avibactam	Ceftibuten	Ceftibuten-ledaborbactam	Meropenem	Meropenem- vaborbactam
CRK0103 (ALRG2756)	OXA-181, NDM-5, SHV-11, TEM-1, CTX-M-15	Lesionc	TD insertionh	WT	>16	>8	>16	>16	>4	>4
CRK0102 (ALRG2757)	OXA-181, SHV-11, CTX-M-15	Lesionc	TD insertionh	WT	>16	0.5	>16	0.25	>4	>4
CRK0297 (ALRG2881)	OXA-48, SHV-11, TEM-1	Lesiond	Intact	WT	0.5	0.25	0.25	0.25	>4	>4
CRK0298 (ALRG2882)	OXA-48, SHV-11, TEM-1,	Lesiond	Intact	WT	0.5	0.25	0.25	0.25	>4	>4
CRK0035 (ALRG2946)	OXA-48, OXA-1, SHV-1, CTX-M-15	Intact	Intact	WT	>16	1	16	0.12	2	2
CRK0299 (ALRG3126)	OXA-232, SHV-1, TEM-1	Lesione	TD insertionh	WT	>16	1	>16	1	>4	>4
CRK0300 (ALRG3143)	OXA-232, SHV-1, TEM-1	Lesione	TD insertionh	WT	>16	1	>16	0.5	>4	>4
CRK0188 (ALRG1995)	KPC-2, SHV-12, TEM-1	Lesionf	GD insertionh	WT	>16	4	>16	1	>4	>4
CRK0089 (ALRG2019)	KPC-3, SHV-12	Lesionf	Intact	F383Y	>16	>8	>16	4	>4	0.5
CRK0154 (ALRG2490)	KPC-2, SHV-11	Lesionf	GD insertionh	WT	>16	2	>16	2	>4	>4
CRK0114 (ALRG2834)	KPC-4, SHV-1, OXA-1, TEM-1	Lesiong	Intact	WT	>16	>8	>16	<=0.12	2	<= 0.03
CRK0113 (ALRG2835)	KPC-new (KPC-4 W105G), SHV-1	Lesiong	Lesioni	WT	>16	>8	>16	>16	>4	>4
CRK0126 (ARLG2563)	KPC-3, SHV-160	Lesionf	Intact	WT	>16	4	16	8	>4	0.06
CRK0249 (ARLG2562)	KPC-3, SHV-12, OXA-9, TEM-1	Lesionf	Lesionj	WT	>16	2	>16	8	>4	1

^a MIC values of β-lactam/β-lactamase inhibitor combinations are shown as MICs of the β-lactam component.

^b PBP3, penicillin-binding protein 3.

^c IS insertion at codon 274 that prevents expression of a functional porin.

^d Frameshift by 11 bp deletion starting at codon 254 that prevents expression of a functional porin.

^e Premature stop at codon 107 that prevents expression of a functional porin.

^f Frameshift by 1 bp insertion at codon 40 that prevents expression of a functional porin.

^g Frameshift by 2 bp insertion at codon 20 that prevents expression of a functional porin.

^h Key di-amino acid insertion in the extracellular loop 3 region that restricts entry of many β-lactams into the bacterial cell (21).

ⁱ IS insertion at codon 82 that prevents expression of a functional porin.

^j Frameshift by 2-bp insertion at codon 12 that prevents expression of a functional porin.

Three studies of the in vitro activity of ceftibuten-ledaborbactam have recently been published. The first study reported on the activity of ceftibuten-ledaborbactam against a 2018–2020 global culture collection of 3,889 clinical isolates of Enterobacterales, including MDR organisms, ESBL- and carbapenemase-positive organisms, and organisms that were non-susceptible to other antimicrobials (7). At ≤1 mg/L, ceftibuten-ledaborbactam inhibited 54.1% of carbapenem-non-susceptible isolates. Against serine carbapenemase positive isolates, ceftibuten-ledaborbactam inhibited 85.9% of KPC-positive (MIC₉₀, 2 mg/L) and 82.9% of OXA-48-group-positive (MIC₉₀, 2 mg/L) isolates; our findings were similar, with 92.5% of serine carbapenemase positive isolates susceptible at the breakpoint used in the reported study with an MIC₉₀ of 1 mg/L.

In the second study, the MIC₉₀ of ceftibuten-ledaborbactam was 2 mg/L against 1,066 MDR urinary isolates of Enterobacterales from a 2014–2016 global culture collection, with 89.1% of all isolates tested inhibited by ≤1 mg/L (3). This potency was similar to that of the parenteral agents, ceftazidime-avibactam (MIC₉₀, 2 mg/L) and meropenem-vaborbactam (MIC₉₀, 1 mg/L). MIC₉₀ value of ceftibuten-ledaborbactam for serine carbapenemase-positive isolates (n = 116) was >32 mg/L with 75.9% susceptible at ≤1 mg/L. The MIC₉₀ of ceftibuten-ledaborbactam against serine carbapenemase-producers in our study was lower (1 mg/L vs. >32 mg/L), and the proportion susceptible at ≤1 mg/L was higher (92.5% vs 75.9%); the reasons for these differences are not clear.

The third study tested 205 Enterobacterales carrying plasmid AmpC (n = 53), ESBL (n = 50), KPC (n = 50), or OXA-48-like (n = 52) encoding genes, with ceftibuten-ledaborbactam MIC₉₀ values of 1, 0.12, 0.5, and 1 mg/L, respectively (22). Ceftibuten-ledaborbactam MIC₉₀ value against KPC producers was 0.5 mg/L, while ceftazidime-avibactam MIC₉₀ was 4 mg/L, very similar to the MIC₉₀ values for these agents in our study (1 mg/L and 4 mg/L, respectively).

In our study, of the 14 isolates producing OXA-48/OXA-48-like β-lactamases or non-susceptible to one or more of the β-lactam/β-lactamase inhibitor agents tested, 13 had lesions causing loss of function of the porin OmpK35 and eight had mutations in the porin gene ompK36. These lesions and mutations are associated with decreased drug permeability (21) (Table 3). While nine of these 14 isolates were non-susceptible to meropenem-vaborbactam, six were non-susceptible to ceftibuten-ledaborbactam. Metallo-β-lactamases are not inhibited by any of the β-lactamase inhibitors included in the β-lactam/β-lactamase inhibitor combinations tested in this study, including ledaborbactam.

Our current study shows that the activity of ceftibuten-ledaborbactam in the context of urinary tract infections is comparable to that of ceftazidime-avibactam, cefepime-taniborbactam, and meropenem-vaborbactam, against carbapenemase-producing and predominantly KPC-producing K. pneumoniae, with MIC_50/90 values of 0.12/1 mg/L and 92.5% and 99.0%, respectively, of isolates inhibited at EUCAST and CLSI breakpoints. With the advantage of being the only oral agent in this group, ceftibuten-ledaborbactam may have potential as an option for oral treatment of UTIs caused by these organisms and further development is needed to assess its clinical utility.