In Vitro Activities of β-Lactam–β-Lactamase Inhibitor Antimicrobial Agents against Cystic Fibrosis Respiratory Pathogens

We tested the in vitro activities of ceftazidime-avibactam, ceftolozane-tazobactam, meropenem-vaborbactam, piperacillin-tazobactam, and 11 other antimicrobial agents against 420 Burkholderia, Achromobacter, Stenotrophomonas, and Pandoraea strains, 89% of which were cultured from respiratory specimens from persons with cystic fibrosis.

ABSTRACT

We tested the in vitro activities of ceftazidime-avibactam, ceftolozane-tazobactam, meropenem-vaborbactam, piperacillin-tazobactam, and 11 other antimicrobial agents against 420 Burkholderia, Achromobacter, Stenotrophomonas, and Pandoraea strains, 89% of which were cultured from respiratory specimens from persons with cystic fibrosis. Among the β-lactam–β-lactamase inhibitor agents, meropenem-vaborbactam had the greatest activity against Burkholderia and Achromobacter, including multidrug-resistant and extensively-drug-resistant strains. None of the newer β-lactam–β-lactamase combination drugs showed increased activity compared to that of the older agents against Stenotrophomonas maltophilia or Pandoraea spp.

INTRODUCTION

Persons with cystic fibrosis (CF) are susceptible to respiratory tract infection with a suite of opportunistic bacterial pathogens that are relatively infrequent causes of infection in healthy hosts. Species within the Burkholderia cepacia complex, Burkholderia gladioli, Stenotrophomonas maltophilia, and certain Achromobacter species can cause chronic infection of CF airways that may be associated with poorer outcomes (1). Pandoraea species are less frequently encountered but are also capable of chronic airway infection in CF (2). A common feature of these species is broad-range resistance to antimicrobial agents, which contributes to the difficulty in effectively managing infection in this patient population. While a robust pipeline exists to develop novel anti-infective therapies for CF airway infection, there is an immediate need to explore the utility of newer commercially available antimicrobials for their activity against this group of CF respiratory pathogens.

The development of new β-lactam–β-lactamase inhibitor agents has been spurred by the increasing prevalence of Gram-negative bacteria carrying transmissible β-lactamases (3). Although antimicrobial resistance in CF respiratory pathogens is attributable to multiple mechanisms, the role of β-lactamases has garnered recent attention, particularly with respect to species in the Burkholderia cepacia complex (4, 5). We tested the activity of ceftazidime-avibactam, ceftolozane-tazobactam, meropenem-vaborbactam, piperacillin-tazobactam, and 11 comparator antimicrobial agents against 420 isolates from the strain collection of the Burkholderia cepacia Research Laboratory and Repository (BcRLR) at the University of Michigan. This set included 200 Burkholderia isolates, 100 Achromobacter isolates, 100 S. maltophilia isolates, and 20 Pandoraea isolates. Isolates were recovered between 2013 and 2018 from 420 persons receiving care in 171 medical centers in 119 cities in 43 U.S. states, plus Toronto, ON, Canada. Most isolates (89%) were cultured from respiratory specimens from persons with CF, and most (88%) were isolated between 2014 and 2017.

All strains had been putatively identified by referring laboratories and sent to the BcRLR for species level confirmation using genetic-based methods previously described (6–9). All Burkholderia strains were distinct by genotypic analyses (10), and strains of all other species were recovered from different patients across a wide geographic range. The distribution of species within each genus roughly reflected that found in CF patients as follows: Burkholderia cenocepacia (50 isolates), Burkholderia multivorans (50 isolates), Burkholderia gladioli (50 isolates), Burkholderia vietnamiensis (20 isolates), Burkholderia cepacia (20 isolates), Burkholderia contaminans (10 isolates), Achromobacter xylosoxidans (50 isolates), Achromobacter ruhlandii (25 isolates), Achromobacter dolens (25 isolates), Pandoraea apista (10 isolates), and Pandoraea sputorum (10 isolates). MIC values were measured using the reference Clinical and Laboratory Standards broth microdilution method (11), with β-lactamase inhibitors added at fixed concentrations. Custom-dried antibiotic plates were read on a Sensititre ARIS instrument (Thermo Fisher Scientific, Waltham, MA, USA).

The results of susceptibility testing comparing ceftazidime-avibactam, ceftolozane-tazobactam, meropenem-vaborbactam, and piperacillin-tazobactam, as well as ceftazidime and meropenem, are shown in Table 1. These agents showed a wide range of activity against Achromobacter spp., Burkholderia spp., and S. maltophilia. Whereas the activity of meropenem-vaborbactam was consistent across all Burkholderia spp. tested, ceftazidime-avibactam and ceftolozane-tazobactam were 4- to 16-fold less active against B. gladioli than against B. cepacia complex spp. Although piperacillin-tazobactam had relatively poor activity against Burkholderia spp. as a group (MIC₉₀ of 32 μg/ml), it showed good activity against B. gladioli and B. vietnamiensis (MIC₉₀ values of ≤2 μg/ml) (see Table S1 in the supplemental material).

TABLE 1.

Activities of β-lactam–β-lactamase inhibitor antimicrobial agents and comparators against bacterial strains

Species or group (no. of isolates)	Antimicrobial agent	MIC (μg/ml)			% Susceptiblea
		Range	MIC50	MIC90
Achromobacter (100)	Ceftazidime	1 to >32	8	32	71
	Ceftazidime-avibactam	1 to >32	8	32	78
	Ceftolozane-tazobactam	≤0.5 to >32	>32	>32	1
	Meropenem	≤0.5 to >32	1	>32	72
	Meropenem-vaborbactam	≤0.5 to 32	≤0.5	8	86
	Piperacillin-tazobactam	≤2 to >128	≤2	128	87
Burkholderia cepacia complex (150)	Ceftazidime	≤0.5 to >32	4	8	91
	Ceftazidime-avibactam	≤0.5 to >32	4	4	97
	Ceftolozane-tazobactam	≤0.5 to >32	1	8	89
	Meropenem	≤0.5 to >32	2	4	90
	Meropenem-vaborbactam	≤0.5 to >32	1	2	97
	Piperacillin-tazobactam	≤2 to >128	4	64	85
Burkholderia gladioli (50)	Ceftazidime	4 to >32	16	32	20
	Ceftazidime-avibactam	2 to >32	16	16	24
	Ceftolozane-tazobactam	2 to >32	16	32	12
	Meropenem	≤0.5 to 4	1	2	100
	Meropenem-vaborbactam	≤0.5 to 4	1	2	100
	Piperacillin-tazobactam	≤2 to 4	≤2	≤2	100
Stenotrophomonas maltophilia (100)	Ceftazidime	≤0.5 to >32	32	>32	34
	Ceftazidime-avibactam	≤0.5 to >32	16	>32	40
	Ceftolozane-tazobactam	≤0.5 to >32	32	>32	27
	Meropenem	≤0.5 to >32	>32	>32	11
	Meropenem-vaborbactam	≤0.5 to >32	>32	>32	12
	Piperacillin-tazobactam	≤2 to >128	128	>128	18
Pandoraea (20)	Ceftazidime	>32	>32	>32	0
	Ceftazidime-avibactam	>32	>32	>32	0
	Ceftolozane-tazobactam	>32	>32	>32	0
	Meropenem	32 to >32	>32	>32	0
	Meropenem-vaborbactam	32 to >32	>32	>32	0
	Piperacillin-tazobactam	8 to >128	64	>128	5

^aSusceptibility based on CLSI breakpoints established for Pseudomonas aeruginosa as follows: ceftazidime, ≤8 μg/ml; ceftazidime-avibactam, ≤8 μg/ml; ceftolozane-tazobactam, ≤4 μg/ml; meropenem, ≤4 μg/ml; meropenem-vaborbactam, ≤4 μg/ml; piperacillin-tazobactam, ≤16 μg/ml.

Meropenem-vaborbactam and piperacillin-tazobactam demonstrated the greatest activity among the β-lactam–β-lactamase inhibitor agents against Achromobacter spp.; a comparable majority of strains would be considered susceptible based on CLSI breakpoints established for Pseudomonas aeruginosa (Table 1; see also Table S2 in the supplemental material). Whereas ceftazidime and ceftazidime-avibactam showed intermediate activity (MIC₅₀ of 8 μg/ml; MIC₉₀ of 32 μg/ml) against Achromobacter spp., ceftolozane-tazobactam showed poor activity (MIC₅₀ of >32 μg/ml). None of the β-lactam–β-lactamase inhibitor agents showed good activity against S. maltophilia or Pandoraea spp. (MIC₉₀ values of ≥32 μg/ml).

Detailed results of susceptibility testing at the species level, along with 11 comparator drugs, are shown in Table S3 in the supplemental material. Ceftazidime-avibactam and ceftazidime had equivalent activities overall against Burkholderia spp. and against Achromobacter spp. An exception was the greater activity of ceftazidime-avibactam versus ceftazidime against B. multivorans (MIC₉₀ values of 4 and 16 μg/ml, respectively). Similarly, the activity of meropenem-vaborbactam was comparable to that of meropenem against Burkholderia spp., while meropenem-vaborbactam showed greater potency than meropenem against Achromobacter spp. (MIC₉₀ values of 8 and >32 μg/ml, respectively).

Among the comparator drugs, trimethoprim-sulfamethoxazole, minocycline, and tigecycline had the greatest activities overall against the species tested, while aztreonam and colistin had generally poor activity. The fluoroquinolone and carbapenem agents showed greater activity against B. gladioli than against the other Burkholderia spp. tested. The fluoroquinolone agents also exhibited good activity (MIC₅₀ values of 4 μg/ml) against Pandoraea strains; however, while imipenem showed good activity (MIC₅₀ of 2 μg/ml), meropenem and meropenem-vaborbactam did not (both with MIC₅₀ values of >32 μg/ml).

As is typical for the genera included in this study, the majority (67%) of isolates tested were multidrug resistant (MDR) and 6% were extensively drug resistant (XDR) based on criteria used to define MDR and XDR in P. aeruginosa (12). All four of the β-lactam–β-lactamase inhibitor drugs showed good activity (MICs of ≤4 μg/ml) against the majority of MDR/XDR Burkholderia spp., while only meropenem-vaborbactam and piperacillin-tazobactam showed good activity against the majority of MDR/XDR Achromobacter spp. (see Fig. S1 in the supplemental material). Meropenem-vaborbactam and ceftazidime-avibactam were the most active against the nine XDR strains of Burkholderia tested, with 67% and 22%, respectively, of strains being inhibited by ≤4 μg/ml of these agents. The activity of all four β-lactam–β-lactamase drugs was poor against most MDR/XDR S. maltophilia strains.

Airway infection in CF typically involves complex polymicrobial bacterial communities that often include more than one opportunistic pathogen (13). As such, in vitro susceptibility testing of a single species isolated from the community performs poorly in predicting clinical outcomes of antimicrobial therapy (14). Despite this limitation, the relative activity of antimicrobial agents against species recovered in culture is an important consideration in guiding antibiotic choice. Clearly, choosing an agent with greater in vitro activity than one with little or no activity against a species known to be associated with poor outcomes in CF (i.e., the species included in this study) remains a goal of therapy.

The in vitro data presented here indicate that β-lactam–β-lactamase inhibitor agents, including the newer agents ceftazidime-avibactam, ceftolozane-tazobactam, and meropenem-vaborbactam, offer therapeutic options for managing airway infections due to opportunistic respiratory pathogens in persons with CF.