In Vitro Activities and Inoculum Effects of Ceftazidime-Avibactam and Aztreonam-Avibactam against Carbapenem-Resistant Enterobacterales Isolates from South Korea

Ceftazidime-avibactam (CAZ-AVI) and aztreonam-avibactam (AZT-AVI) are novel antibiotic combinations active against multidrug-resistant Gram-negative pathogens. This study aimed to evaluate their in vitro activities and inoculum effects in carbapenem-resistant Enterobacterales (CRE), including carbapenemase-producing (CP)-CRE and non-CP-CRE. A total of 81 independent clinical isolates of carbapenem-resistant Escherichia coli and Klebsiella pneumoniae were collected. CAZ-AVI and AZT-AVI minimal inhibitory concentrations (MICs) were evaluated by broth microdilution using standard and high inocula. The inoculum effect was defined as an ≥8-fold increase in MIC with high inoculum. Phenotypic determination of β-lactam resistance mechanism and PCR for carbapenemase genes were performed. Of the 81 CRE isolates, 35 (43%) were CP-CRE. Overall, 73% of the isolates were susceptible to CAZ-AVI, and 95% had low AZT-AVI MICs (≤8 µg/mL). The MIC50/MIC90s of CAZ-AVI and AZT-AVI were 4/≥512 µg/mL and 0.5/4 µg/mL, respectively. CAZ-AVI was more active against non-CP-CRE than against CP-CRE (susceptibility 80% vs. 63%, p = 0.08; MIC50/MIC90, 2/16 μg/mL vs. 4/≥512 μg/mL), whereas AZT-AVI was more active against CP-CRE (MIC50/MIC90, 0.25/1 μg/mL vs. 0.5/8 μg/mL). All four isolates with high AZT-AVI MIC (≥16 μg/mL) were resistant to CAZ-AVI, but only 18% (4/22) of CAZ-AVI-resistant isolates had high AZT-AVI MIC. The rates of the inoculum effect for CAZ-AVI and AZT-AVI were 18% and 47%, respectively (p < 0.001). Interestingly, the frequency of the AZT-AVI inoculum effect was higher in K. pneumoniae than E. coli (64% vs. 8%, p < 0.001). AZT-AVI is more active against CRE than CAZ-AVI, even in CP-CRE and CAZ-AVI-resistant isolates. The presence of a substantial inoculum effect may contribute to clinical failure in high-inoculum infections treated with AZT-AVI.


Introduction
Carbapenem-resistant Enterobacterales (CRE) challenge pharmaceutical chemists and clinicians on account of their difficult-to-treat resistance and increasing global prevalence [1]. Due to the limited therapeutic options for CRE infections, the polymyxins (colistin and polymyxin B) are frequently used as last resort drugs. However, their high rates of nephrotoxicity, which range from 30% to 60%, make their use problematic [2]. Avibactam, a new non-β-lactam β-lactamase inhibitor, is an inhibitor of class A β-lactamases, including extended-spectrum β-lactamases (ESBLs) and Klebsiella pneumoniae carbapenemases (KPCs), and also class C (AmpC) and some class D (OXA-48) β-lactamases [3]. Ceftazidime-avibactam shows promising activity against CRE strains, such as KPC-producing K. pneumoniae and Escherichia coli [4]. However, ceftazidime-avibactam is usually not active against class B metallo-β-lactamase (MBL)-producing CRE [4]. After the introduction of ceftazidime-avibactam into clinical use, cases of resistance due to various mechanisms have been increasingly reported [5].
Aztreonam, a monobactam, is unique among currently used β-lactams, in that it is stable to hydrolysis by MBLs [6]. However, it is easily inactivated by ESBLs, AmpC, and KPCs. When combined with avibactam, aztreonam can inhibit cell wall synthesis in MBL-producing bacteria, despite the presence of co-carried β-lactamases such as ESBLs and AmpC [7]. Thus, aztreonam-avibactam can be more effective than ceftazidime-avibactam against MBL-producing strains. However, there are limited data on the susceptibility of aztreonam-avibactam to CRE. In addition, β-lactam antibiotics, especially β-lactam/β-lactamase inhibitors, are known to display an inoculum effect of variable severity against Gram-negative bacteria. The inoculum effect is a laboratory phenomenon described as a significant increase in the minimal inhibitory concentration (MIC) of an antibiotic when the number of bacteria inoculated increases [8]. Reduced in vitro activity of the β-lactams against a dense bacterial population is commonly attributed to the presence of a high level of β-lactamase. In a clinical situation involving a high bacterial burden, such as an abscess, the inoculum effect may lead to treatment failure.
This study aimed to compare the in vitro activities of ceftazidime-avibactam and aztreonam-avibactam and their inoculum effects in carbapenemase-producing (CP-CRE) and non-CP-CRE isolates. It also examined the relationship of their in vitro activities and inoculum effects to the carbapenem resistance mechanisms of the target bacteria.

Bacterial Isolates and Study Design
Carbapenem-resistant E. coli and K. pneumoniae isolates were collected from consecutive patients who had no prior exposure to ceftazidime-avibactam or aztreonam-avibactam at Asan Medical Center, a 2700-bed tertiary care center in Seoul, South Korea. A total of 81 carbapenem-resistant E. coli and K. pneumoniae isolates (25 and 56 isolates, respectively) were collected from January 2014 to October 2018. Only the first CRE isolate from each patient was included in the study. Species identification and initial antimicrobial susceptibility testing was performed with a MicroScan Walk-Away plus System using a Neg Combo Panel Type 72 (Dade Behring Inc., West Sacramento, CA, USA). CRE isolates, defined as those resistant to meropenem or imipenem (MIC ≥ 4 µg/mL) [9,10], were stored at −80 • C. They were streaked on an agar plates and incubated 24 h before experiments. The composition of the study isolates by specimen source was as follows: blood stream, n = 57 (70.4% of all isolates tested), intra-abdominal, n = 9 (11.1%), urinary tract, n = 7 (8.6%), respiratory tract, n = 5 (6.2%), and soft tissue, n = 3 (3.7%).

Antimicrobial Susceptibility Testing and the Inoculum Effect
Antimicrobial susceptibility testing for ceftazidime, aztreonam, ceftazidime-avibactam, aztreonam-avibactam, meropenem, colistin, and tigecycline was performed in triplicate using standard broth microdilution [9,11]. Avibactam was tested at a fixed concentration of 4 µg/mL. MICs were Antibiotics 2020, 9, 912 3 of 12 interpreted according to the Clinical and Laboratory Standards Institute (CLSI) breakpoints for all antimicrobial agents except for those for which CLSI breakpoints are not yet available [9]: aztreonam-avibactam, for which clinical breakpoints have not yet been assigned and tigecycline and colistin, for which the European Committee on Antimicrobial Susceptibility Testing (EUCAST) MIC breakpoints were applied [12]. To determine whether there was an inoculum effect with ceftazidime-avibactam, aztreonam-avibactam, and meropenem, the MICs of each antibiotic with high inocula (1 × 10 7 CFU/mL) were compared to those with standard inocula (1 × 10 5 CFU/mL) [13,14]. Our definition of an inoculum effect was an 8-fold or greater MIC increase in testing with the high inoculum [8,15]. As MIC values > 256 µg/mL for ceftazidime-avibactam were not further examined, the presence of the inoculum effect for such isolates was not determined. E. coli ATCC 25922 and K. pneumoniae ATCC 700603 were used as quality control strains for each test. All results determined with these strains were within the CLSI quality control ranges. Ceftazidime, aztreonam, meropenem, tigecycline, and colistin were purchased from Sigma-Aldrich (St. Louis, MO, USA) and avibactam was obtained from AdooQ Bioscience (Irvine, CA, USA).

Basis of Resistance and Molecular Identification of β-Lactamase Genes
The carbapenem resistance mechanisms of each isolate were examined to determine the impact of resistance mechanisms on antimicrobial susceptibility patterns and the inoculum effect. The modified carbapenem inactivation method, with high sensitivity and specificity, was used to confirm carbapenemase production in all the study CRE isolates [16]. For the carbapenemase-producing (CP) isolates, PCR was used to amplify carbapenemase genes (bla KPC , bla IMP , bla VIM , bla NDM , and bla OXA-48-like) according to the procedures described in previous studies [17][18][19]. Non-carbapenemase-producing (non-CP) isolates usually acquire carbapenem resistance by the production of extended-spectrum β-lactamase (ESBL) and/or AmpC cephalosporinase (AmpC) in conjunction with membrane impermeability or active drug efflux. To identify the β-lactamase types among the non-CP-CRE, the presence of ESBL was determined by the MicroScan ESBL detection test (included in Neg Combo Panel Type 72) using cefotaxime and ceftazidime alone and in combination with clavulanic acid. For isolates not confirmed by the MicroScan ESBL detection test, the double-disk synergy test was performed in addition, using cefotaxime (30 µg), ceftazidime (30 µg), cefepime (30 µg), and amoxicillin plus clavulanate (20 µg and 10 µg each) disks [20,21]. As non-susceptibility to cefoxitin (MIC >8 µg/mL) is considered a surrogate marker for the presence of high-level production of AmpC, isolates non-susceptible to cefoxitin were further characterized by the AmpC confirmatory test using cefoxitin and cloxacillin [22,23]. Cefepime, ceftazidime, cefotaxime, and amoxicillin-clavulanic acid disk were purchased from Bio-rad (Hercules, CA, USA), and cefoxitin disks were obtained from Oxoid (Basingstoke, UK).

Statistical Analysis
Differences between groups were analyzed using the χ 2 test or Fisher's exact test as appropriate. A two-sided p < 0.05 was considered statistically significant. SPSS version 24.0 (IBM, Armonk, NY, USA) was used in the statistical analyses.

Discussion
In the present study, ceftazidime-avibactam was active against 73% of CRE isolates, and aztreonam-avibactam had a low MIC (≤8 µg/mL) against 95% of the CRE isolates. In total, 43% of the study isolates were CP-CRE isolates, of which 34% harbored MBL. Unlike ceftazidime-avibactam, aztreonam-avibactam was less active against non-CP-CRE isolates than against CP-CRE isolates. The inoculum effect was more consistently detected with aztreonam-avibactam than with ceftazidime-avibactam, especially in K. pneumoniae isolates. To our knowledge, this is the first study to compare the in vitro activities of ceftazidime-avibactam and aztreonam-avibactam together with evaluating the inoculum effect against CRE isolates encompassing CP-CRE and non-CP CRE.
Our data suggest that aztreonam-avibactam may be more active than ceftazidime-avibactam against CRE strains. This result is similar to previous reports that found lower MIC 50 /MIC 90 values for aztreonam-avibactam than for ceftazidime-avibactam against Gram-negative bacilli [24][25][26][27][28][29][30]. However, those studies included only CP-CRE, either alone or along with non-CRE. To date, ceftazidime-avibactam stands out as one of the most important additions to the antimicrobial armamentarium, as it is the first marketed fixed combination with activity against CRE, including those with the OXA and KPC enzymes [4,31]. Notably, avibactam cannot inhibit MBL, and nor can any other new β-lactamase inhibitor such as vaborbactam and relebactam [29,32]. Aztreonam, a monobactam, is stable to MBL [33]. In previous studies, aztreonam, in combination with ceftazidime-avibactam or avibactam, showed promising activity against MBL in Enterobacterales [7, 34,35]. Since the Food and Drug Administration (FDA) granted Qualified Infectious Disease Product and Fast Track designation to aztreonam-avibactam for CRE infections in November 2019, further efforts to place aztreonam-avibactam in the right position to combat against CRE are critical. Recently, a Chinese study assessed the in vitro activities of ceftazidime-avibactam and aztreonam-avibactam against 58 CRE isolates, including both CP-CRE and non-CP-CRE [36]. In that study, the non-CP-CRE (n = 14) had lower aztreonam-avibactam MICs than the CP-CRE. However, in our hands, aztreonam-avibactam was less active against non-CP-CRE (n = 46), than CP-CRE. In addition, an aztreonam-avibactam inoculum effect was more common in the non-CP-CRE. Some differences in the species composition of study isolates and/or molecular epidemiology may have caused different results between the two studies. Further study of more CRE isolates is needed to confirm these findings. Our data showed that the K. pneumoniae isolates were more resistant to colistin or tigecycline than the E. coli isolates, but they tended to have lower MICs for both ceftazidime-avibactam and aztreonam-avibactam. A between-species comparison of antimicrobial activity could also be helpful in the management of CRE infection.
In this study, aztreonam-avibactam had a higher rate of inoculum effects than ceftazidime-avibactam (47% and 18%, respectively), particularly in K. pneumoniae isolates. Given the growing body of concern over the high mortality and rapid dissemination of CRE infection, it is surprising that no studies have explored the inoculum effects of ceftazidime-avibactam and aztreonam-avibactam-the two essential therapeutic options-in CRE isolates. Whether the inoculum effect is clinically significant remains debatable [8,37,38]. In the era of carbapenem-resistance driven in large by a broader β-lactamase repertoire, the bacterial inoculum can reduce the activity of antimicrobial agents, particularly for β-lactam/β-lactamase combination drugs. Based on our data, aztreonam-avibactam may fail in the treatment of high-inoculum infections caused by CRE. Considering this, the susceptibility breakpoint for aztreonam-avibactam against CRE should be set at <8 µg/mL, although more clinical evidence is needed.
Our study has several limitations. First, the possibility of resistance mechanisms other than carbapenemase co-existing in the CPE isolates were not examined. Thus, we did not assess the entire resistance mechanism-specific impact on antimicrobial susceptibility in an ideal manner. Second, of the members of the Enterobacterales, we selectively collected E. coli and K. pneumoniae species, and the species mainly harboring intrinsic AmpC β-lactamases were not included. Further studies of these species are warranted. Additionally, ceftazidime-avibactam and aztreonam-avibactam have not been in clinical use in South Korea so the absence of prior exposure to these antimicrobial agents may have affected the MIC values of these two agents and resistance rates. Despite these limitations, the in vitro activities of ceftazidime-avibactam and aztreonam-avibactam against the CRE isolates in our study are consistent with previous CP-CRE reports based on extensive epidemiologic data. Moreover, this study provided data for non-CP-CRE, for which there has been less reliable clinical data than for CP-CRE. Future studies to determine the optimal dosing and breakpoints of aztreonam-avibactam, and the benefit of combination therapy are warranted.
Ceftazidime-avibactam has been considered a reasonable option for the treatment of CRE infection. In this study, aztreonam-avibactam was more active against CRE than ceftazidime-avibactam, even for CP-CRE and ceftazidime-avibactam-resistant isolates. Currently available data may render aztreonam-avibactam a "game changer" in the treatment of difficult-to-treat Gram-negative organisms of various resistance mechanisms, including MBL. However, aztreonam-avibactam is not a one-size-fits-all option. The presence of the substantial inoculum effect may contribute to clinical failure in patients treated with aztreonam-avibactam for high inoculum CRE infections.