Activity of Ceftazidime-Avibactam against Extended-Spectrum- and AmpC β-Lactamase-Producing Enterobacteriaceae Collected in the INFORM Global Surveillance Study from 2012 to 2014

Abstract

The in vitro activity of ceftazidime-avibactam was evaluated against 34,062 isolates of Enterobacteriaceae from patients with intra-abdominal, urinary tract, skin and soft-tissue, lower respiratory tract, and blood infections collected in the INFORM (International Network For Optimal Resistance Monitoring) global surveillance study (176 medical center laboratories in 39 countries) in 2012 to 2014. Overall, 99.5% of Enterobacteriaceae isolates were susceptible to ceftazidime-avibactam using FDA approved breakpoints (susceptible MIC of ≤8 μg/ml; resistant MIC of ≥16 μg/ml). For individual species of the Enterobacteriaceae, the ceftazidime-avibactam MIC inhibiting ≥90% of isolates (MIC₉₀) ranged from 0.06 μg/ml for Proteus species to 1 μg/ml for Enterobacter spp. and Klebsiella pneumoniae. Carbapenem-susceptible isolates of Escherichia coli, K. pneumoniae, Klebsiella oxytoca, and Proteus mirabilis with a confirmed extended-spectrum β-lactamase (ESBL) phenotype, or a ceftazidime MIC of ≥16 μg/ml if the ESBL phenotype was not confirmed by clavulanic acid inhibition, were characterized further to identify the presence of specific ESBL- and plasmid-mediated AmpC β-lactamase genes using a microarray-based assay and additional PCR assays. Ceftazidime-avibactam demonstrated potent activity against molecularly confirmed ESBL-producing (n = 5,354; MIC₉₀, 0.5 μg/ml; 99.9% susceptible), plasmid-mediated AmpC-producing (n = 246; MIC₉₀, 0.5 μg/ml; 100% susceptible), and ESBL- and AmpC-producing (n = 152; MIC₉₀, 1 μg/ml; 100% susceptible) isolates of E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis. We conclude that ceftazidime-avibactam demonstrates potent in vitro activity against globally collected clinical isolates of Enterobacteriaceae, including isolates producing ESBLs and AmpC β-lactamases.

INTRODUCTION

Ceftazidime is an extended-spectrum, parenteral cephalosporin that originally was approved for use in patients in the 1980s (http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm). At that time it demonstrated broad-spectrum activity against Gram-positive cocci and Gram-negative bacilli, including Pseudomonas aeruginosa. The introduction of extended-spectrum cephalosporins (e.g., ceftazidime, cefotaxime, and ceftriaxone) into clinical use was soon followed by the identification of Enterobacteriaceae and other Gram-negative bacilli that harbored extended-spectrum β-lactamases (ESBLs) that hydrolyzed these extended-spectrum cephalosporins, thereby conferring resistance to these agents. Today, hundreds of genetically unique ESBLs have been identified, characterized, and catalogued (1, 2). Even more recently, plasmid-mediated β-lactamases (i.e., AmpC β-lactamases, Klebsiella pneumoniae carbapenemases [KPC], and metallo-β-lactamases) also have appeared in clinical isolates of Enterobacteriaceae that inactivate cephalosporins as well as other β-lactams (1, 2).

One approach to address the increasing prevalence and broadening spectrum of β-lactamases is to generate improved β-lactamase inhibitors. Older β-lactamase inhibitors include clavulanate, sulbactam, and tazobactam in combination with amoxicillin or ticarcillin, ampicillin or cefoperazone, and piperacillin, respectively; these combination agents are active only against Gram-negative bacilli harboring Ambler class A enzymes, including some isolates of Enterobacteriaceae that produce ESBLs. Avibactam is a novel, diazobicyclooctane, non-β-lactam β-lactamase inhibitor that protects β-lactams against hydrolysis in Gram-negative bacteria that produce Ambler class A enzymes, including ESBLs and serine carbapenemases, such as KPCs, class C enzymes (e.g., AmpC β-lactamases such as CMY-2), and some class D enzymes (e.g., OXA-48) (3–6). Avibactam inactivates β-lactamases with an active-site serine via acylation (opening of the avibactam ring). When the unbound avibactam is removed, deacylation of β-lactamases generally occurs very slowly, releasing intact, nonhydrolyzed avibactam (4, 7) as opposed to hydrolysis characteristic of inactivation with clavulanic acid, tazobactam, and sulbactam. In February 2015, ceftazidime-avibactam was approved by the U.S. Food and Drug Administration for the treatment of complicated intra-abdominal infections (in combination with metronidazole) and complicated urinary tract infections (including pyelonephritis) in adults who have limited or no alternative treatment options (8). Clinical trials involving ceftazidime-avibactam are under way for the indication of hospital-acquired bacterial pneumonia/ventilator-associated bacterial pneumonia.

The INFORM (International Network For Optimal Resistance Monitoring) global surveillance program was established to benchmark and track the in vitro activity of ceftazidime-avibactam and comparative agents against clinical isolates, including β-lactamase-producing Enterobacteriaceae, before and after regulatory approval in multiple international markets. This report summarizes results from the INFORM global surveillance study conducted in 2012 to 2014, focusing primarily on the activity of ceftazidime-avibactam against Enterobacteriaceae with ESBLs and AmpC β-lactamases.

MATERIALS AND METHODS

Bacterial isolates.

The 2012-2014 INFORM global surveillance program included medical center laboratories in Europe (19 countries, 93 laboratories), the Asia/Pacific region (9 countries, 41 laboratories), Latin America (6 countries, 26 laboratories), and the Middle East/Africa (5 countries, 16 laboratories). The INFORM global surveillance program requests participating medical center laboratories to annually collect predefined numbers of isolates of selected bacterial species from patients with specific types of infections (one isolate per patient infectious episode). In total, 34,062 isolates of Enterobacteriaceae were collected from patients with intra-abdominal, urinary tract, skin and soft-tissue, lower respiratory tract, and blood infections. All study isolates collected by medical center laboratories were shipped to a central reference laboratory, International Health Management Associates, Inc. (IHMA; Schaumburg, IL, USA), where isolate identities were confirmed using a Bruker Biotyper MALDI-TOF (matrix-assisted laser desorption ionization-time of flight) instrument (Bruker Daltonics, Billerica, MA, USA).

Antimicrobial susceptibility testing.

Antimicrobial susceptibility testing of ceftazidime-avibactam and a range of comparative agents was performed by IHMA, following CLSI methodology, using in-house-prepared, 96-well broth microdilution panels that were stored at −80°C and thawed before use (9, 10). Avibactam was tested at a fixed concentration of 4 μg/ml, in combination with doubling dilutions of ceftazidime from 0.015 to 128 μg/ml. MICs were interpreted using CLSI breakpoints (9). In the absence of CLSI breakpoints, ceftazidime-avibactam and tigecycline MICs were interpreted using FDA MIC interpretive breakpoints (ceftazidime-avibactam, ≤8 μg/ml [susceptible] and ≥16 μg/ml [resistant]; tigecycline, ≤2 μg/ml [susceptible], 4 μg/ml [intermediate], and ≥8 μg/ml [resistant]) (8, 11). For colistin, MICs were interpreted using EUCAST interpretative breakpoints (≤2 μg/ml [susceptible] and ≥4 μg/ml [resistant]) (12).

Screening isolates of Escherichia coli, K. pneumoniae, Klebsiella oxytoca, and Proteus mirabilis for ESBL and plasmid-mediated AmpC β-lactamase genes.

Isolates of E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis that were susceptible to carbapenems (ertapenem, meropenem, doripenem, and imipenem) were screened for possible ESBL and/or plasmidic AmpC production using ceftazidime or aztreonam MICs (>1 μg/ml) and were confirmed as ESBL producers using combination clavulanic acid-based testing by following CLSI guidelines (9). Molecular analysis was performed on all E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis isolates phenotypically confirmed as ESBL producers, as well as on all carbapenem-susceptible E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis isolates with ceftazidime MICs of ≥16 μg/ml (ceftazidime resistant) that were insensitive to clavulanic acid inhibition. All identified isolates (n = 5,752) were tested for the presence of β-lactamase genes using a combination of the microarray-based Check-MDR CT101 kit (Check-Points, Wageningen, Netherlands) and published multiplex PCR assays to identify genes encoding ESBLs (TEM, SHV, CTX-M, VEB, PER, and GES) and original-spectrum β-lactamases (OSBLs; TEM and SHV that did not contain substitutions at amino acid positions 104, 164, or 238 [TEM] or 146, 238, or 240 [SHV], which are associated with ESBL activity), as well as plasmid-mediated AmpC β-lactamases (ACC, ACT, CMY, DHA, FOX, MIR, and MOX) and carbapenemases (KPC, OXA-48, IMP, VIM, NDM, SPM, and GIM) as previously described (13). For all molecular assays, genomic DNA was extracted from colonies grown overnight on blood agar (Remel, Lenexa, KS) using the QIAamp DNA minikit and the QIAcube instrument (Qiagen, Valencia, CA) according to the manufacturer's instructions. Enzyme variants were identified as necessary by DNA sequencing and comparison to the National Center for Biotechnology Information database (www.ncbi.nlm.nih.gov) and the Lahey Clinic website (www.lahey.org/studies). The few carbapenem-susceptible isolates that tested positive for genes encoding carbapenemases were removed from further analysis. The identification and characterization of carbapenem-nonsusceptible isolates of Enterobacteriaceae also was excluded from the current analysis and will be presented in a subsequent publication.

RESULTS

Ceftazidime-avibactam inhibited 99.5% (33,877 of 34,062 isolates) of Enterobacteriaceae isolates tested (Table 1) at a ceftazidime concentration of ≤8 μg/ml, the FDA-approved breakpoint for susceptibility. In comparison, percentages of susceptibility for other agents were lower against all isolates of Enterobacteriaceae, with >90% of isolates testing as susceptible to ertapenem, meropenem, amikacin, and tigecycline while <80% of isolates were susceptible to ceftazidime, cefepime, aztreonam, and levofloxacin. A few isolates (n = 185) were nonsusceptible to ceftazidime-avibactam, and 177 of them (95.7%) also were carbapenem nonsusceptible. These 185 isolates were distributed among 58 sites in 26 countries and comprised 14 species of Enterobacteriaceae (data not shown). The majority (n = 144; 77.8%) of these were metallo-β-lactamase producers (NDM, IMP, or VIM), β-lactamases that are not inhibited by avibactam.

TABLE 1.

In vitro activities of ceftazidime-avibactam and comparative antimicrobial agents tested against Enterobacteriaceae isolates collected by the INFORM global surveillance program in 2012 to 2014

Organism (no.) and antimicrobial agent	MIC (μg/ml)			MIC interpretationa (%)
	50%	90%	Range	Susceptible	Intermediate	Resistant
Enterobacteriaceae (34,062)
Ceftazidime-avibactam	0.12	0.5	≤0.015–>128	99.5		0.5
Ceftazidime	0.25	64	≤0.015–>128	75.6	1.9	22.5
Cefepimeb	≤0.12	>16	≤0.12–>16	77.3	4.9	17.9
Aztreonam	0.12	64	≤0.015–>128	74.3	1.9	23.8
Piperacillin-tazobactam	2	128	≤0.25–>128	84.0	5.9	10.1
Ertapenemc (n = 20,885)	0.015	0.25	≤0.002–>1	94.5	1.6	3.9
Meropenem	0.03	0.12	≤0.004–>8	97.2	0.5	2.3
Amikacin	2	8	≤0.25–>32	96.3	1.5	2.2
Colistind	≤0.12	>4	≤0.12–>4	83.1		16.9
Tigecycline	0.5	2	≤0.015–>8	92.9	5.7	1.4
Levofloxacin	0.06	>4	≤0.03–>4	75.0	2.5	22.5
E. coli (11,770)
Ceftazidime-avibactam	0.12	0.25	≤0.015–>128	99.9		0.1
Ceftazidime	0.25	32	≤0.015–>128	78.9	2.6	18.5
Cefepime	≤0.12	>16	≤0.12–>16	76.1	4.4	19.5
Aztreonam	0.12	64	≤0.015–>128	75.6	2.9	21.5
Piperacillin-tazobactam	2	32	≤0.25–>128	89.8	4.8	5.4
Ertapenem (n = 7,505)	0.015	0.12	≤0.002–>1	98.7	0.5	0.8
Meropenem	0.03	0.03	≤0.004–>8	99.6	0.1	0.3
Amikacin	4	8	≤0.25–>32	97.8	1.1	1.1
Colistin	≤0.12	≤0.12	≤0.12–>4	99.7		0.3
Tigecycline	0.25	0.5	≤0.015–8	99.8	0.2	0
Levofloxacin	0.12	>4	≤0.03–>4	63.1	1.5	35.4
K. pneumoniae (9,098)
Ceftazidime-avibactam	0.12	1	≤0.015–>128	99.0		1.0
Ceftazidime	0.5	>128	≤0.015–>128	60.3	1.7	38.0
Cefepime	≤0.12	>16	≤0.12–>16	61.4	6.2	32.4
Aztreonam	0.12	>128	≤0.015–>128	60.2	1.0	38.8
Piperacillin-tazobactam	4	>128	≤0.25–>128	71.2	8.5	20.3
Ertapenem (n = 5,034)	0.03	>1	≤0.002–>1	87.4	2.4	10.2
Meropenem	0.03	0.25	≤0.004–>8	91.6	1.2	7.2
Amikacin	2	16	≤0.25–>32	93.0	3.1	3.9
Colistin	≤0.12	0.25	≤0.12–>4	97.2		2.8
Tigecycline	0.5	2	≤0.015–>8	96.1	3.3	0.6
Levofloxacin	0.12	>4	≤0.03–>4	70.8	3.2	26.0
Enterobacter spp.e (3,931)
Ceftazidime-avibactam	0.25	1	≤0.015–>128	98.8		1.2
Ceftazidime	0.5	128	≤0.015–>128	66.4	1.8	31.8
Cefepime	≤0.12	8	≤0.12–>16	83.8	6.8	9.4
Aztreonam	0.12	64	≤0.015–>128	67.6	2.4	30.0
Piperacillin-tazobactam	4	128	≤0.25–>128	74.9	11.3	13.8
Ertapenem (n = 2,492)	0.06	1	≤0.002–>1	86.5	6.1	7.4
Meropenem	0.06	0.12	≤0.004–>8	97.8	0.6	1.6
Amikacin	2	4	≤0.25–>32	97.8	0.9	1.3
Colistin	≤0.12	0.25	≤0.12–>4	95.0		5.0
Tigecycline	0.5	2	≤0.015–8	96.6	2.8	0.6
Levofloxacin	0.06	2	≤0.03–>4	90.4	2.0	7.6
P. mirabilis (2,235)
Ceftazidime-avibactam	0.06	0.06	≤0.015–>128	99.8		0.2
Ceftazidime	0.06	2	≤0.015–>128	94.4	1.7	3.9
Cefepime	≤0.12	8	≤0.12–>16	87.3	3.9	8.8
Aztreonam	≤0.015	1	≤0.015–>128	95.4	1.1	3.5
Piperacillin-tazobactam	0.5	2	≤0.25–>128	98.4	0.7	0.9
Ertapenem (n = 1,449)	0.015	0.06	0.004–>1	99.3	0.2	0.5
Meropenem	0.06	0.12	≤0.004–>8	99.4	0.2	0.4
Amikacin	4	8	≤0.25–>32	94.5	1.1	4.4
Colistin	>4	>4	≤0.12–>4	0.9		99.1
Tigecycline	4	8	0.03–>8	41.2	44.8	14.0
Levofloxacin	0.12	>4	≤0.03–>4	78.5	4.5	17.0
Citrobacter spp.f (1,889)
Ceftazidime-avibactam	0.12	0.5	≤0.015–>128	99.3		0.7
Ceftazidime	0.25	128	≤0.015–>128	78.6	1.1	20.4
Cefepime	≤0.12	2	≤0.12–>16	91.3	3.5	5.2
Aztreonam	0.12	32	≤0.015–>128	77.9	1.8	20.3
Piperacillin-tazobactam	2	64	≤0.25–>128	84.5	7.9	7.6
Ertapenem (n = 1,209)	0.015	0.25	≤0.002–>1	96.9	1.2	1.9
Meropenem	0.03	0.06	≤0.004–>8	99.0	0.2	0.8
Amikacin	2	4	≤0.25–>32	97.8	0.4	1.8
Colistin	≤0.12	0.25	≤0.12–>4	99.6		0.4
Tigecycline	0.5	1	≤0.015–4	99.0	1.0	0
Levofloxacin	0.06	2	≤0.03–>4	90.8	2.3	6.9
K. oxytoca (1,900)
Ceftazidime-avibactam	0.12	0.25	≤0.015–>128	99.5		0.5
Ceftazidime	0.12	1	≤0.015–>128	93.7	1.1	5.2
Cefepime	≤0.12	2	≤0.12–>16	92.4	4.4	3.2
Aztreonam	0.12	32	≤0.015–>128	85.6	1.4	13.0
Piperacillin-tazobactam	2	128	≤0.25–>128	87.7	1.2	11.1
Ertapenem (n = 1,194)	0.015	0.06	≤0.002–>1	98.2	0.5	1.3
Meropenem	0.03	0.06	≤0.004–>8	98.9	0.3	0.8
Amikacin	2	4	≤0.25–>32	98.9	0.4	0.7
Colistin	≤0.12	0.12	≤0.12–>4	99.9		0.1
Tigecycline	0.5	1	≤0.015–4	99.2	0.8	0
Levofloxacin	0.06	0.5	≤0.03–>4	95.5	1.6	2.9
M. morganii (979)
Ceftazidime-avibactam	0.06	0.12	≤0.015–16	99.9		0.1
Ceftazidime	0.12	8	≤0.015–>128	88.0	2.7	9.3
Cefepime	≤0.12	0.25	≤0.12–>16	94.6	2.4	3.0
Aztreonam	0.03	2	≤0.015–>128	95.1	1.8	3.1
Piperacillin-tazobactam	0.5	2	≤0.25–>128	98.1	0.9	1.0
Ertapenem (n = 641)	0.03	0.06	0.008–>1	99.8	0	0.2
Meropenem	0.12	0.25	0.008–1	100	0	0
Amikacin	2	8	≤0.25–>32	98.4	0.7	0.9
Colistin	>4	>4	≤0.12–>4	1.1		98.9
Tigecycline	2	4	0.12–>8	77.0	17.6	5.4
Levofloxacin	0.06	>4	≤0.03–>4	81.7	6.1	12.2
P. vulgaris (995)
Ceftazidime-avibactam	0.06	0.06	≤0.015–32	99.9		0.1
Ceftazidime	0.06	0.12	≤0.015–>128	98.4	0.2	1.4
Cefepime	≤0.12	0.25	≤0.12–>16	97.6	1.5	0.9
Aztreonam	≤0.015	0.12	≤0.015–>128	98.5	0.2	1.3
Piperacillin-tazobactam	≤0.25	1	≤0.25–>128	99.7	0.1	0.2
Ertapenem (n = 607)	0.03	0.06	0.004–>1	99.8	0	0.2
Meropenem	0.06	0.12	≤0.004–4	99.9	0	0.1
Amikacin	2	4	≤0.25–>32	99.2	0.2	0.6
Colistin	>4	>4	≤0.12–>4	1.5		98.5
Tigecycline	2	4	≤0.015–>8	81.3	16.2	2.5
Levofloxacin	0.06	0.5	≤0.03–>4	96.2	2.4	1.4
S. marcescens (785)
Ceftazidime-avibactam	0.25	0.5	≤0.015–>128	99.2		0.8
Ceftazidime	0.25	4	0.03–>128	91.0	1.0	8.0
Cefepime	≤0.12	2	≤0.12–>16	90.7	2.2	7.1
Aztreonam	0.12	16	≤0.015–>128	88.2	1.1	10.7
Piperacillin-tazobactam	2	16	≤0.25–>128	93.4	3.6	3.0
Ertapenem (n = 488)	0.03	0.25	≤0.002–>1	98.0	0.2	1.8
Meropenem	0.06	0.12	0.008–>8	98.8	0	1.2
Amikacin	2	8	≤0.25–>32	94.8	1.3	3.9
Colistin	>4	>4	≤0.12–>4	7.8		92.2
Tigecycline	1	2	0.12–>8	94.0	5.1	0.9
Levofloxacin	0.12	2	≤0.03–>4	93.2	2.6	4.2
Providencia spp.g (316)
Ceftazidime-avibactam	0.12	0.5	≤0.015–>128	97.5		2.5
Ceftazidime	0.12	16	≤0.015–>128	88.3	1.6	10.1
Cefepime	≤0.12	8	≤0.12–>16	88.0	4.1	7.9
Aztreonam	≤0.015	4	≤0.015–>128	92.1	1.6	6.3
Piperacillin-tazobactam	1	8	≤0.25–>128	96.2	1.3	2.5
Ertapenem (n = 188)	0.03	0.12	0.008–1	99.5	0.5	0
Meropenem	0.06	0.25	≤0.004–>8	98.4	1.0	0.6
Amikacin	2	8	≤0.25–>32	95.9	0.9	3.2
Colistin	>4	>4	≤0.12–>4	1.6		98.4
Tigecycline	2	4	0.25–>8	66.5	27.5	6.0
Levofloxacin	0.5	>4	≤0.03–>4	59.5	9.8	30.7
Othersh (164)
Ceftazidime-avibactam	0.12	0.25	≤0.015–1	100		0
Ceftazidime	0.12	0.5	≤0.015–>128	98.2	0.6	1.2
Cefepime	≤0.12	0.25	≤0.12–>16	98.2	0	1.8
Aztreonam	0.06	0.25	≤0.015–>128	97.6	0	2.4
Piperacillin-tazobactam	2	4	≤0.25–>128	99.4	0	0.6
Ertapenem (n = 78)	0.015	0.06	0.004–0.25	100	0	0
Meropenem	0.06	0.12	0.015–>8	99.4	0	0.6
Amikacin	1	4	≤0.25–>32	98.8	0	1.2
Colistin	≤0.12	>4	≤0.12–>4	57.3		42.7
Tigecycline	0.5	2	0.06–>8	91.5	7.9	0.6
Levofloxacin	0.06	0.25	≤0.03–4	97.6	2.4	0

^aMICs were interpreted according to CLSI breakpoints (9), with the exception of ceftazidime-avibactam, where MICs were interpreted using the MIC interpretative criteria according to the FDA (10), tigecycline, where MICs were interpreted using the MIC interpretative criteria according to the FDA 11), and colistin, where EUCAST breakpoints were applied (12).

^bFor cefepime, the intermediate category was replaced by susceptible-dose dependent (SDD) in 2014.

^cErtapenem was not tested against isolates collected in 2014.

^dColistin was tested with 0.002% polysorbate-80.

^eThe 3,931 isolates of Enterobacter spp. were composed of E. cloacae (2,207 isolates), E. aerogenes (1,350), E. asburiae (250), E. kobei (93), E. ludwigii (24), E. gergoviae (4), E. amnigenus (1), E. cancerogenus (1), and E. hormaechei (1).

^fThe 1,889 isolates of Citrobacter spp. were composed of C. freundii (1,033 isolates), C. koseri (707), C. braakii (109), C. amalonaticus (21), C. murliniae (5), nonspeciated Citrobacter spp. (4), C. farmeri (4), C. sedlakii (3), C. gillenii (1), C. diversus (1), and C. youngae (1).

^gThe 316 isolates of Providencia spp. were composed of P. stuartii (161 isolates), P. rettgeri (141), and P. alcalifaciens (14).

^hThe others category was composed of species not included in other subsets: Escherichia fergusonii (1 isolate), Escherichia hermannii (2), Escherichia vulneris (1), Hafnia alvei (4), Klebsiella ozaenae (1), Klebsiella variicola (7), Kluyvera ascorbata (2), Pantoea agglomerans (2), Proteus hauseri (3), Proteus penneri (42), Proteus rettgeri (2), Raoultella ornithinolytica (57), Raoultella planticola (13), Raoultella terrigena (2), Serratia liquefaciens (16), Serratia odorifera (1), Serratia rubidaea (1), and Serratia ureilytica (7).

The overall MIC₉₀ for ceftazidime-avibactam for Enterobacteriaceae was 0.5 μg/ml, which was 128-fold lower than the MIC₉₀ for ceftazidime alone (64 μg/ml) (Table 1). Ceftazidime-avibactam MIC₉₀ values for individual species of Enterobacteriaceae were 0.06 μg/ml for P. mirabilis (n = 2,235) and Proteus vulgaris (n = 995); 0.12 μg/ml for Morganella morganii (n = 979); 0.25 μg/ml for E. coli (n = 11,770) and K. oxytoca (n = 1,900); 0.5 μg/ml for Citrobacter spp. (n = 1,889), Serratia marcescens (n = 785), and Providencia spp. (n = 316); and 1 μg/ml for Enterobacter spp. (n = 3,931) and K. pneumoniae (n = 9,098). The greatest difference in percent susceptibility for ceftazidime-avibactam and ceftazidime for individual species of Enterobacteriaceae was observed for K. pneumoniae (38.7% difference) and Enterobacter spp. (32.4% difference).

Table 2 summarizes the in vitro activities of ceftazidime-avibactam and comparative antimicrobial agents against the 5,752 isolates of E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis identified as harboring an ESBL, a plasmid-mediated AmpC β-lactamase, or both.

TABLE 2.

In vitro activities of ceftazidime-avibactam and comparative antimicrobial agents against carbapenem-susceptible E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis isolates harboring an ESBL, a plasmid-mediated AmpC β-lactamase, or both^a

Organism (no.) and Antimicrobial agent	MIC(μg/ml)			MIC interpretationb(%)
	50%	90%	Range	Susceptible	Intermediate	Resistant
All (5,752)
Ceftazidime-avibactam	0.25	0.5	≤0.015–16	99.9		0.1
Ceftazidime	32	128	0.12–>128	16.1	6.1	77.8
Cefepimec	>16	>16	≤0.12–>16	10.9	16.2	72.9
Aztreonam	64	128	0.03–>128	7.9	6.6	85.5
Piperacillin-tazobactam	16	>128	≤0.25–>128	67.7	16.0	16.3
Ertapenemd (n = 3,307)	0.06	0.5	≤0.002–0.5	100	0	0
Meropenem	0.03	0.06	≤0.004–1	100	0	0
Amikacin	4	16	≤0.25–>32	93.3	2.5	4.2
Colistine	≤0.12	0.12	≤0.015–>4	98.3		1.7
Tigecycline	0.5	2	≤0.015–>8	97.4	2.1	0.5
Levofloxacin	>4	>4	≤0.03–>4	34.4	4.3	61.3
ESBL-positive with/without OSBLf (5,354)
Ceftazidime-avibactam	0.25	0.5	≤0.015–16	99.9		0.1
Ceftazidime	32	128	0.12–>128	17.1	6.4	76.5
Cefepime	>16	>16	≤0.12–>16	6.9	16.8	76.3
Aztreonam	64	128	0.03–>128	7.4	5.4	87.2
Piperacillin-tazobactam	16	>128	≤0.25–>128	68.0	16.1	15.9
Ertapenem (n = 3,062)	0.06	0.5	≤0.002–0.5	100	0	0
Meropenem	0.03	0.06	≤0.004–1	100	0	0
Amikacin	4	16	≤0.25–>32	93.5	2.5	4.0
Colistin	≤0.12	0.12	≤0.015–>4	98.4		1.6
Tigecycline	0.5	2	≤0.015–8	97.4	2.2	0.4
Levofloxacin	>4	>4	≤0.03–>4	34.1	4.5	61.4
AmpC-positive with/without OSBL (246)
Ceftazidime-avibactam	0.12	0.5	≤0.015–8	100		0
Ceftazidime	32	>128	2–>128	1.6	1.6	96.8
Cefepime	0.25	2	≤0.12–>16	91.0	5.3	3.7
Aztreonam	8	32	0.12–>128	20.7	35	44.3
Piperacillin-tazobactam	8	>128	1–>128	67.5	14.2	18.3
Ertapenem (n = 158)	0.12	0.5	0.008–0.5	100	0	0
Meropenem	0.03	0.06	≤0.004–0.25	100	0	0
Amikacin	2	8	0.5–>32	95.9	0	4.1
Colistin	≤0.12	0.25	≤0.015–>4	97.6		2.4
Tigecycline	0.25	1	0.06–>8	96.8	2.0	1.2
Levofloxacin	>4	>4	≤0.03–>4	45.1	1.6	53.3
ESBL-positive plus AmpC-positive, with/without OSBL (152)
Ceftazidime-avibactam	0.25	1	≤0.015–8	100		0
Ceftazidime	64	>128	1–>128	2.0	3.3	94.7
Cefepime	>16	>16	≤0.12–>16	19.1	13.1	67.8
Aztreonam	64	>128	2–>128	3.3	3.3	93.4
Piperacillin-tazobactam	16	>128	0.5–>128	57.2	13.2	29.6
Ertapenem (n = 87)	0.25	0.5	0.015–0.5	100	0	0
Meropenem	0.06	0.12	0.015–1	100	0	0
Amikacin	4	>32	0.5–>32	82.2	4.0	13.8
Colistin	≤0.12	0.25	≤0.015–>4	96.1		3.9
Tigecycline	0.5	1	0.06–>8	96.7	2.0	1.3
Levofloxacin	>4	>4	0.06–>4	27.0	2.6	70.4

^aAll 5,752 isolates were susceptible to carbapenems (ertapenem, doripenem, imipenem, and meropenem).

^bMICs were interpreted according to CLSI breakpoints (9), with the exception of ceftazidime-avibactam, where MICs were interpreted using the MIC interpretative criteria according to the FDA (10), tigecycline, where MICs were interpreted using the MIC interpretative criteria according to the FDA (11), and colistin, where EUCAST breakpoints were applied (12).

^cFor cefepime, the intermediate category was replaced by susceptible-dose dependent (SDD) in 2014.

^dErtapenem was not tested against isolates collected in 2014.

^eColistin was tested with 0.002% polysorbate-80.

^fOSBL, original-spectrum β-lactamase (e.g., TEM-1, TEM-2, SHV-1, SHV-11).

Subset analysis of 5,354 ESBL-producing isolates, some of which also harbored an original-spectrum β-lactamase (e.g., TEM-1, TEM-2, SHV-1, and SHV-11), included 2,457 isolates (92.4% of 2,658 isolates) from Europe, 1,148 isolates (84.0% of 1,366 isolates) from the Asia/Pacific region, 1,057 isolates (92.9% of 1,138 isolates) from Latin America, and 692 isolates (93.0% of 744 isolates) from the Middle East/Africa. AmpC enzymes were identified in all regions; 120, 72, 36, and 18 isolates, respectively, from the Asia/Pacific region, Europe, Latin America, and the Middle East/Africa. Isolates in which genes for both AmpC and ESBL enzymes were detected also were detected in all regions, with the majority of isolates (81/152; 53.3%) collected from the Asia/Pacific region.

ESBL-positive isolates (99.9% susceptible; MIC₉₀, 0.5 μg/ml), AmpC-positive isolates (100% susceptible; MIC₉₀, 0.5 μg/ml), and isolates positive for both an ESBL and an AmpC β-lactamase (100% susceptible; MIC₉₀, 1 μg/ml) were almost uniformly susceptible to ceftazidime-avibactam (Table 2). In contrast, the activities of piperacillin-tazobactam (68.0%, 67.5%, and 57.2% susceptible), cefepime (6.9%, 91.0%, and 19.1% susceptible), and ceftazidime (17.1%, 1.6%, and 2.0% susceptible) were substantially impacted by the presence of ESBLs, AmpC β-lactamases, or both types of β-lactamases, respectively.

Table 3 provides MIC distributions for ceftazidime-avibactam for isolates grouped by the presence of one or more ESBLs, plasmidic AmpC β-lactamases, or both enzyme types in isolates of carbapenem-susceptible E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis. Of the 5,354 ESBL-positive isolates, 5,060 (94.5%) harbored a single ESBL. Specific ceftazidime-avibactam MIC distributions are provided for frequently cited ESBLs (CTX-M-15, CTX-M-14, SHV-12, VEB, GES, and PER). Ceftazidime-avibactam activity was not affected by the presence of more than one ESBL or by a specific ESBL. Among the plasmid-mediated AmpC enzymes identified, CMY-2 and DHA-1 were the most common (87.0%; 214/246). Isolates containing these enzymes showed MIC₉₀ values of ≤0.5 μg/ml, which were lower than those containing other plasmidic AmpC enzymes (MIC₉₀, 2 μg/ml). The 152 isolates that possessed both an ESBL and AmpC β-lactamase showed a MIC₉₀ value of 1 μg/ml. Only eight isolates, each from a different country, tested with a ceftazidime-avibactam MIC of 8 μg/ml (i.e., at the susceptible breakpoint). These isolates carried CTX-M-15 only (two isolates), CTX-M-27 only (one), PER-3 and TEM-1 (one), PER-type and TEM-1 (one), CMY-131 and TEM-36 (one), CTX-M-15 and CMY-42 (one), and SHV-11, TEM-1, CTX-M-15, and ACC-1 (one). Only one isolate identified in the ESBL group resulted in a ceftazidime-avibactam MIC of 16 μg/ml; this isolate was a K. oxytoca strain collected in Europe, carried SHV-12 and TEM-1, and was carbapenem (meropenem), amikacin, colistin, levofloxacin, and tigecycline susceptible, with a piperacillin-tazobactam MIC of >128 μg/ml (resistant). The β-lactamases that were identified in this isolate are susceptible to inhibition by avibactam; therefore, the mechanism by which this isolate is resistant has yet to be determined.

TABLE 3.

MIC frequency and distributions for ceftazidime-avibactam against carbapenem-susceptible E. coli, K. pneumoniae, K. oxytoca, and P. mirabilis isolates harboring an ESBL, a plasmid-mediated AmpC β-lactamase, or both^a

β-Lactamase gene(s) presentc (no. of isolates)	No. (cumulative %) of isolates inhibited at ceftazidime-avibactam MIC (μg/ml) ofb:
	≤0.015	0.03	0.06	0.12	0.25	0.5	1	2	4	8	16
All ESBL-positive with/without OSBLd (5,354)	71 (1.3)	242 (5.9)	569 (16.5)	1607 (46.5)	1554 (75.5)	893 (92.2)	331 (98.4)	58 (99.5)	23 (99.9)	5 (99.9)	1 (100)
≥2 ESBLs present with/without OSBL (294)	4 (1.4)	10 (4.8)	21 (11.9)	50 (28.9)	82 (56.8)	71 (81.0)	46 (96.6)	8 (99.3)	2 (100)
CTX-M-15 only with/without OSBL (3,487)	32 (0.9)	155 (5.4)	306 (14.1)	988 (42.5)	1092 (73.8)	629 (91.8)	231 (98.5)	36 (99.5)	16 (99.9)	2 (100)
CTX-M-14 only with/without OSBL (420)	9 (2.1)	18 (6.4)	81 (25.7)	187 (70.2)	90 (91.7)	33 (99.5)	2 (100)
SHV-12 only with/without OSBL (171)	4 (2.3)	6 (5.8)	16 (15.2)	36 (36.3)	56 (69.0)	37 (90.6)	9 (95.9)	4 (98.2)	2 (99.4)	0 (99.4)	1 (100)
VEB only with/without OSBLe (2)						1 (50.0)	1 (50.0)
GES only with/without OSBLe (1)							1 (100)
PER only with/without OSBLe (3)				1 (33.3)	0 (33.3)	0 (33.3)	0 (33.3)	0 (33.3)	0 (33.3)	2 (100)
Other ESBL only with/without OSBLf (976)	22 (2.3)	53 (7.7)	145 (22.5)	345 (57.9)	234 (81.9)	122 (94.4)	41 (98.6)	10 (99.6)	3 (99.9)	1 (100)
All AmpC-positive with/without OSBLg (246)	2 (0.8)	9 (4.5)	38 (19.9)	74 (50.0)	76 (80.9)	32 (93.9)	8 (97.2)	3 (98.4)	3 (99.6)	1 (100)
CMY-2 only with/without OSBL (159)	2 (1.3)	5 (4.4)	28 (22.0)	58 (58.5)	51 (90.6)	12 (98.1)	2 (99.4)	0 (99.4)	1 (100)
DHA-1 only with/without OSBL (54)		4 (7.4)	7 (20.4)	11 (40.7)	17 (72.2)	11 (92.6)	3 (98.1)	1 (100)
Other AmpC only with/without OSBLh (32)			3 (9.4)	4 (21.9)	8 (46.9)	9 (75.0)	3 (84.4)	2 (90.6)	2 (96.9)	1 (100)
All ESBL-positive and AmpC-positive with/without OSBL (152)	1 (0.7)	10 (7.2)	13 (15.8)	22 (30.3)	49 (62.5)	29 (81.6)	19 (94.1)	4 (96.7)	3 (98.7)	2 (100)

^aAll isolates were susceptible to carbapenems (ertapenem, doripenem, imipenem, and meropenem).

^bThe MIC₉₀ is boldfaced for each MIC distribution.

^cOSBL, original-spectrum β-lactamase (e.g., TEM-1, TEM-2, SHV-1, and SHV-11).

^dAll ESBL-positive isolates, including those with multiple β-lactamase enzymes.

^eToo few isolates to calculate the MIC₉₀.

^fIsolates carrying a single ESBL other than CTX-M-15, CTX-M-14, SHV-12, VEB, PER, or GES.

^gAll AmpC-positive isolates, including those with multiple β-lactamase enzymes. One isolate carried both CMY-2 and DHA-1 and was not included in the subset analyses.

^hIsolates carrying a single AmpC other than CMY-2 or DHA-1.

DISCUSSION

ESBLs, plasmid-mediated AmpC β-lactamases, and carbapenemases pervade clinical isolates of Enterobacteriaceae worldwide (14–22). Geographic differences in β-lactamase composition and prevalence have been reported, require ongoing monitoring, and will continue to evolve over time under the influences of antimicrobial selective pressure and international travel (5, 14–22). Global surveillance studies consistently identify CTX-M-14, CTX-M-15, and SHV-12 to be the most prevalent ESBLs and CMY-2 as the most common plasmid-mediated AmpC β-lactamase currently circulating (2, 3, 15).

The emergence and spread of antimicrobial-resistant bacteria is often the result of one or more successful clones, such as E. coli sequence type 131, and such clones are frequently associated with a multidrug resistance phenotype (23). The results in this study confirmed in vitro and in vivo studies of Enterobacteriaceae and P. aeruginosa that reported that ceftazidime-avibactam overcomes resistance attributable to ESBLs, plasmid-mediated AmpC β-lactamases, and carbapenemases with serine active sites (e.g., KPC), including multidrug-resistant isolates, but does not inhibit the growth of isolates producing metallo-β-lactamases (5, 6, 18–21, 24–27). In the current study, 99.5% of all Enterobacteriaceae isolates tested were susceptible to ceftazidime-avibactam (Table 1), including the vast majority of ceftazidime-resistant and ESBL- and AmpC β-lactamase-positive isolates. Ceftazidime-avibactam appeared equally active against isolates with ESBLs, AmpC β-lactamases, and both types of β-lactamase (Table 2).

In addition, these results confirm previous publications reporting that ceftazidime-avibactam had similar activity (MIC₉₀s ranging from 0.25 to 0.5 μg/ml) against isolates with CTX-M-15-like, CTX-M-14-like, SHV, and CMY-2-like enzymes, as well as against ESBL-negative isolates (19). The current study found similar activity for ceftazidime-avibactam against specific enzyme types, such as CTX-M, SHV-12, other ESBLs, CMY-2, and DHA-1 alone and in combination (Table 3).

Avibactam has been reported to protect ceftazidime from hydrolysis by β-lactamases found in Enterobacteriaceae isolates and P. aeruginosa, but the combination is only moderately active against Acinetobacter spp., anaerobes (excluding Bacteroides fragilis, Clostridium perfringens, Prevotella spp., and Porphyromonas spp.), and Gram-positive cocci (24). It has been hypothesized that some species of Enterobacteriaceae, such as Klebsiella and Enterobacter, have occasional isolates that are not susceptible to ceftazidime-avibactam because they are impermeable to ceftazidime, harbor multiple β-lactamases, including a carbapenemase, and/or hyperproduce AmpC β-lactamase to a level that overwhelms avibactam's inhibitory activity (24). However, in a 2011 study, Livermore and coworkers demonstrated that avibactam reduced the MIC of ceftazidime against 15 isolates of Enterobacter spp. and K. pneumoniae with porin loss in combination with an ESBL or AmpC β-lactamase (5). Lahiri and coworkers have identified residues in AmpC variants with the ability to generate resistance to avibactam, although the authors suggest that the ability of avibactam to mimic the key interaction of a β-lactam substrate combined with its tight binding likely confer a barrier to the development of clinical resistance (28). It is unclear whether any of the mechanisms discussed above played a role in the reduced susceptibilities for the nine carbapenem-susceptible isolates that showed a ceftazidime-avibactam MIC of 8 μg/ml and higher, because the β-lactamases that were identified in these were predicted to be inhibitable by avibactam. These mechanisms also may play a role in the ceftazidime-avibactam resistance found with 41 isolates (0.1%) that did not contain a metallo-β-lactamase.

In conclusion, we found that ceftazidime-avibactam demonstrated potent in vitro activity against a recent collection of more than 34,000 isolates of Enterobacteriaceae, including ESBL- and AmpC-producing isolates, collected from medical centers in 39 countries. Avibactam broadened the spectrum of activity of ceftazidime to include ESBL- and AmpC β-lactamase-producing Enterobacteriaceae. Avibactam is the first marketed β-lactamase inhibitor with activity against AmpC-mediated resistance; clavulanic acid, sulbactam, and tazobactam only inhibit class A β-lactamase enzymes. Ceftazidime-avibactam appears to be a valuable addition to the limited armamentarium currently available to treat serious Gram-negative infections; however, its activity will depend upon the prevalence of specific resistance mechanisms (e.g., metallo-β-lactamases) circulating in the specific geographic locations where it will be used. Ongoing surveillance for β-lactamase-mediated resistance and other mechanisms of resistance is imperative to ensure the long-term availability and effectiveness of antibacterial agents.

Funding Statement

This investigation was funded by AstraZeneca Pharmaceuticals as part of the sponsored INFORM global surveillance program. The sponsor approved the overall study design. All investigative sites were recruited and study supplies were provided by IHMA, Inc. Analysis of the final MIC and molecular data was performed by IHMA.