In vitro antimicrobial susceptibility of clinical respiratory isolates to ceftazidime-avibactam and comparators (2016–2018)

Background This antimicrobial surveillance study reports in vitro antimicrobial activity and susceptibility data for a panel of agents against respiratory isolates of Enterobacterales and Pseudomonas aeruginosa. Methods Isolates from respiratory specimens were collected in Africa/Middle East, Asia/South Pacific, Europe and Latin America between 2016 and 2018, as part of the Antimicrobial Testing Leadership and Surveillance (ATLAS) program. Broth microdilution methodology was used to quantify minimum inhibitory concentrations, from which rates of susceptibility were determined using EUCAST breakpoints (version 10). Rates of subsets with genes encoding β-lactamases (extended-spectrum β-lactamases [ESBLs], serine carbapenemases and metallo-β-lactamases [MBLs]) were also determined, as well as rates of multidrug-resistant (MDR) P. aeruginosa. Results Among all respiratory Enterobacterales isolates, susceptibility to ceftazidime-avibactam, meropenem, colistin and amikacin was ≥94.4% in each region. For Enterobacterales isolates that were ESBL-positive or carbapenemase-positive/MBL-negative, ceftazidime-avibactam susceptibility was 93.6 and 98.9%, respectively. Fewer than 42.7% of MBL-positive Enterobacterales isolates were susceptible to any agents, except colistin (89.0% susceptible). Tigecycline susceptibility was ≥90.0% among Citrobacter koseri and Escherichia coli isolates, including all β-lactamase-positive subsets. ESBL-positive Enterobacterales were more commonly identified in each region than isolates that were ESBL/carbapenemase-positive; carbapenemase-positive/MBL-negative; or MBL-positive. Among all respiratory P. aeruginosa isolates, the combined susceptibility rates (susceptible at standard dosing regimen plus susceptible at increased exposure) were highest to ceftazidime-avibactam, colistin and amikacin (≥82.4% in each region). Susceptibility to colistin was ≥98.1% for all β-lactamase-positive subsets of P. aeruginosa. The lowest rates of antimicrobial susceptibility were observed among MBL-positive isolates of P. aeruginosa (≤56.6%), with the exception of colistin (100% susceptible). MDR P. aeruginosa were most frequently identified in each region (18.7–28.7%), compared with the subsets of ESBL-positive; carbapenemase-positive/MBL-negative; or MBL-positive isolates. Conclusions Rates of susceptibility among the collections of respiratory Enterobacterales and P. aeruginosa isolates were highest to ceftazidime-avibactam, colistin and amikacin in each region. Tigecycline was active against all subsets of C. koseri and E. coli, and colistin was active against all subsets of P. aeruginosa. The findings of this study indicate the need for continued antimicrobial surveillance among respiratory Gram-negative pathogens, in particular those with genes encoding MBLs. Supplementary Information The online version contains supplementary material available at 10.1186/s12879-021-06153-0.


Introduction
Antimicrobial resistance among Gram-negative bacteria is a long-standing problem that needs to be monitored in an effort to preserve the efficacy of current antimicrobial agents. Pseudomonas aeruginosa and species of Enterobacterales are common causative pathogens of respiratory infections such as hospital-acquired bacterial pneumonia (HABP) and ventilator-associated bacterial pneumonia (VABP), and respiratory infections caused by antimicrobialresistant Enterobacterales and P. aeruginosa are associated with higher patient mortality [1,2]. Among isolates of Enterobacterales and P. aeruginosa, antimicrobial resistance can be mediated by the production of β-lactamases, for example, extended-spectrum β-lactamases (ESBLs), serine carbapenemases and metallo-β-lactamases (MBLs) [3].
Ceftazidime-avibactam is a combination antimicrobial agent comprising ceftazidime, a third-generation cephalosporin, and avibactam, a non-β-lactam β-lactamaseinhibitor. Avibactam inhibits Ambler class A, class C, and certain class D OXA-type β-lactamases, but not MBLs. Therefore ceftazidime-avibactam is active against ESBL-and serine carbapenemase-positive Gramnegative isolates, but not against MBL-positive isolates [4][5][6]. Ceftazidime-avibactam has been approved for the treatment of HABP and VABP, as well as complicated intra-abdominal infection (in combination with metronidazole) and complicated urinary tract infection (including pyelonephritis) [7,8]. In Europe, ceftazidimeavibactam is also indicated for the treatment of infections due to aerobic Gram-negative organisms in adult patients with limited treatment options [8].
The aim of this study is to report in vitro antimicrobial activity and susceptibility data for a panel of antimicrobial agents against isolates of Enterobacterales and P. aeruginosa collected from respiratory specimens as part of the Antimicrobial Testing Leadership and Surveillance (ATLAS) program (2016)(2017)(2018). Data on rates of resistant isolates will also be presented. The geographical regions of collection included in the analysis are Africa/Middle East, Asia/South Pacific, Europe and Latin America.

Methods
Isolates from respiratory specimens were collected from hospitalised patients from participating study centers between 2016 and 2018 in four geographical regions (Africa/Middle East, Asia/South Pacific, Europe and Latin America). Only non-duplicate isolates of the organism considered to be the potential causative pathogen of the infection were included in the study. Demographic information recorded for each isolate included specimen source, patient age and sex, and type of hospital setting.
Isolates were collected and identified at the participating center, and pure cultures were shipped to the central laboratory (International Health Management Associates [IHMA] Inc., Schaumburg, IL, USA). The central laboratory then re-identified and confirmed bacterial species using matrix-assisted laser desorption/ionization timeof-flight mass spectrometry (Bruker Biotyper; Bruker Daltonics, Billerica, MA, USA), and performed antimicrobial susceptibility testing using self-manufactured frozen broth microdilution panels [9]. Ceftazidimeavibactam was tested with a fixed concentration of 4 mg/L avibactam with doubling dilutions of ceftazidime. All minimum inhibitory concentrations (MICs) were interpreted using version 10.0 of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) clinical breakpoint tables [10]. Among the Enterobacterales, isolates of Morganella spp., Proteus spp., Providencia spp. or Serratia spp. were excluded from the analysis for colistin activity, due to intrinsic resistance. All isolates of Enterobacterales were included in the analysis for tigecycline activity, but tigecycline EUCAST breakpoints are only available for isolates of Citrobacter koseri and Escherichia coli. For isolates of P. aeruginosa, EUCAST have revised the susceptible breakpoints at the standard dosing regimen (S) for piperacillin-tazobactam, aztreonam, ceftazidime, cefepime, imipenem and levofloxacin [10], therefore isolates tested against these antimicrobial agents are categorized as either susceptible at increased exposure (I), or as resistant (R). For all antimicrobial agents tested against P. aeruginosa, the combined susceptibility rates (susceptible at standard dosing regimen plus susceptible at increased exposure) are presented here.

Isolates collected
A total of 15,460 isolates from respiratory specimens (10,128 Enterobacterales and 5332 P. aeruginosa) were collected from a total of 51 countries in four geographical regions between 2016 and 2018. The number of centers in each participating country and the number of isolates collected by each center are presented in Supplementary Table 1. More than half of isolates were collected in Europe (58.6%; n = 9055), followed by Asia/ South Pacific (19.7%; n = 3040); Latin America (13.3%; n = 2061) and Africa/Middle East (8.4%; n = 1304).
The majority of isolates from respiratory specimens were collected from male patients (64.5%; n = 9972) and half were from patients aged 65 years and older (50.0%; n = 7723). The percentage of isolates collected from respiratory specimens on general wards (47.9%; n = 7398) was similar to intensive care units (43.4%; n = 6716).

Enterobacterales
Antimicrobial activity and susceptibility data for a panel of antimicrobial agents against the collection of Enterobacterales isolates from respiratory sources are shown by region in Table 1. In each region, the highest rates of susceptibility among the collection of Enterobacterales were to ceftazidime-avibactam, meropenem, colistin and amikacin (≥94.4%). Tigecycline susceptibility in all regions was ≥97.9% among isolates of C. koseri and E. coli, the only species within the Enterobacterales collection to which EUCAST breakpoints apply.
By region, the 2016-2018 rate of ESBL-positive Enterobacterales was lowest in Europe and highest in Africa/ Middle East (Fig. 1). Fewer than 5% of Enterobacterales isolates collected in each region were ESBL/carbapenemase-positive, carbapenemase-positive/MBL-negative or MBL-positive.

Pseudomonas aeruginosa
Antimicrobial activity and susceptibility data for a panel of antimicrobial agents against the collection of P. aeruginosa isolates from respiratory sources are shown by region in Table 3. The combined susceptibility (susceptible at standard dosing regimen plus susceptible at increased exposure) among the collection of P. aeruginosa in each region was highest to colistin, ceftazidimeavibactam and amikacin (≥82.4%). Rates of susceptibility to ceftazidime-avibactam, meropenem and amikacin were lower in Latin America than in Africa/Middle East, Asia/South Pacific and Europe.
No isolates of carbapenemase-positive/MBL-negative P. aeruginosa were susceptible to imipenem. All MBLpositive P. aeruginosa isolates were susceptible to colistin, with ≤19.1% susceptible to amikacin or ceftazidimeavibactam. By region, the 2016-2018 rate of MDR P. aeruginosa was lowest in Asia/South Pacific and highest in Latin America (Fig. 2). The 2016-2018 rate of carbapenemase-positive/MBL-negative P. aeruginosa in each region was ≤3%, with only one such isolate collected in Africa/Middle East and three isolates in Asia/ South Pacific. The 2016-2018 rate of MBL-positive P. aeruginosa was lowest in Asia/South Pacific and highest in Latin America.

Discussion
This study presents in vitro antimicrobial activity and susceptibility data for a panel of antimicrobial agents against respiratory isolates of Enterobacterales and P. aeruginosa, collected as part of the ATLAS program (2016-2018), as well as data on subsets of resistant isolates.
Among the Enterobacterales isolates, rates of antimicrobial susceptibility to amikacin, ceftazidimeavibactam, colistin, meropenem and tigecycline were similar, irrespective of the geographical region of collection. For P. aeruginosa isolates, however, rates of susceptibility to ceftazidime-avibactam, colistin and amikacin were lower in Latin America, when compared with the other regions. In a phase 3 trial of hospitalized adults with HABP or VABP due to Gram-negative pathogens, the overall ceftazidime-avibactam MIC 90 against isolates of Enterobacterales (n = 317) was 0.5 mg/L, and against isolates of P. aeruginosa (n = 101) was 8 mg/L [13]. These data are comparable with the MIC 90 values for ceftazidime-avibactam in the current study, with the exception of isolates from Latin America, where the ceftazidime-avibactam MIC 90 against the collection of P. aeruginosa was 32 mg/L.
The rates of all subsets of resistant P. aeruginosa isolates presented here were higher in Latin America than the other regions, which may explain the lower ceftazidime-avibactam activity and susceptibility seen in Latin America. The rate of MDR P. aeruginosa in Latin America (28.7%) was similar to a 2015-2016 global antimicrobial surveillance study (34.6%), where the rate in Latin America was the highest among seven geographical regions [14]. The 2015-2016 global study included aztreonam and colistin in their MDR study definition [14], whereas the current study definition of MDR P. aeruginosa omitted aztreonam and colistin, based on guidance from the Belgian High Council of Health [15] and the combinations of pathogens and antimicrobial agents under continued European Antimicrobial Resistance Surveillance Network (EARS-Net) surveillance [16]. A 2012-2015 study of clinical P. aeruginosa isolates from Latin America reported that 35.8% (643/1794) of P. aeruginosa isolates were meropenem-nonsusceptible (% intermediate plus % resistant) [17]; similar to the current study (37.0% [298/805]).
In the present study, susceptibility to ceftazidimeavibactam among carbapenemase-positive/MBL-negative Enterobacterales isolates remained high (98.9%). This was comparable to the 2012-2015 rate of ceftazidimeavibactam susceptibility reported among isolates of Enterobacterales, collected from the same regions as the current study, that were OXA-48-positive and MBLnegative (99.2%) [6]. The lack of ceftazidime-avibactam activity against MBL-positive isolates (due to the hydrolysis of both ceftazidime and avibactam by the MBL class of β-lactamases [18]) is clearly demonstrated in the present study. In addition, ceftazidime-avibactam susceptibility was notably lower among ESBL/carbapenemase-positive Enterobacterales (64.7%), compared with ESBL-positive isolates (93.6%). A total of 124 isolates in the ESBL/carbapenemase-positive subset were ceftazidime-avibactam-resistant, of which 123 were  MBL-positive and the remaining isolate was carbapenemase-positive/MBL-negative (data not shown).
The single carbapenemase-positive/MBL-negative isolate had a ceftazidime-avibactam MIC of 32 mg/L (EUCAST resistance breakpoint, > 8 mg/L). Isolates of Enterobacterales have been found to coproduce MBLs and Ambler class A β-lactamases, such as ESBLs [19]. The limitations of this study were the predefined number of isolates per species, as well as the variability in center and country participation between the study years, meaning that these results cannot be interpreted as epidemiology findings. The details on the type or size of study centers are not recorded at the time of isolate collection, which could limit the clinical relevance of the data. Despite this, the data reported here highlight the presence of antimicrobialresistant respiratory Gram-negative pathogens in the four geographical regions presented. Ceftazidimeavibactam was active against resistant isolates of Enterobacterales and P. aeruginosa, with the exception of those organisms that were MBL-positive; among this subset of Enterobacterales, susceptibility was highest to colistin. This study provides valuable information to clinicians on the susceptibility of these resistant isolates to antimicrobial agents in current use. Continued monitoring of the antimicrobial susceptibility profiles of respiratory Enterobacterales and P. aeruginosa isolates is necessary to identify the most difficult-to-treat respiratory pathogens and improve patient outcomes.   For the majority of agents, antimicrobial susceptibility was reduced among the subsets of resistant Gramnegative isolates; most notably the MBL-positive subset. Monitoring of antimicrobial susceptibility is therefore Isolates in this subset are carbapenemase-positive and MBL-positive necessary to help physicians to choose appropriate treatment options and improve the treatment outcomes for respiratory Gram-negative infections.