Introduction

Aztreonam-avibactam is currently under clinical investigation for treatment of Gram-negative infections, including those caused by Enterobacterales producing MBLs and/or serine carbapenemases (CPEs) [1]. Aztreonam was approved by the FDA in 1986, and it still is the only clinically available member of the monobactam class [2]. Aztreonam is stable to hydrolysis by MBLs; however, it is hydrolysed by most clinically relevant serine β-lactamases, such as ESBLs, AmpC, and KPC. Because Enterobacterales isolates that produce an MBL usually coproduce a serine β-lactamase, aztreonam was combined with avibactam. Avibactam is a non-β-lactam β-lactamase inhibitor that inhibits the activities of Ambler class A (including extended-spectrum β-lactamases), class C, and some class D β-lactamases [1].

Enterobacterales can express a broad range of mechanisms of antimicrobial resistance, and the treatment of infections caused by multidrug-resistant (MDR) Enterobacterales, especially carbapenem-resistant Enterobacterales (CRE), remains an important challenge for physicians. Resistance to carbapenems in Enterobacterales is usually due to the acquisition of CPEs or overexpressed cephalosporinases combined with decreased permeability. Although globally distributed in many Enterobacterales species, certain CPEs are associated with specific regions or countries [3,4,5]. KPC-producing Enterobacterales, mainly K. pneumoniae, have been extensively reported in the USA and some European countries, such as Greece and Italy [6]. OXA-48 and its derivatives (e.g., OXA-181 and OXA-232) hydrolyse narrow-spectrum β-lactams and weakly hydrolyse carbapenems, but spare broad-spectrum cephalosporins, such as ceftazidime and cefepime. OXA-48-producing Enterobacterales are endemic in Turkey and are frequently reported in several European countries, such as France and Belgium [7]. Class B, or MBLs, are commonly identified in Enterobacterales and Pseudomonas aeruginosa in some geographic regions. NDM, VIM, and IMP are the most frequent MBLs identified in Enterobacterales worldwide. NDM-producing Enterobacterales have been identified globally, with the highest prevalence in the Indian subcontinent, the Middle East, and southeast Europe. VIM-producing Enterobacterales are common in Italy and Greece, whereas IMP is mainly found in Acinetobacter baumannii from China, Japan, or Australia [8, 9].

In the present study, we assessed the in vitro activity of aztreonam-avibactam against a large collection of contemporary (2019–2020) clinical Enterobacterales isolates recovered from patients hospitalised in European medical centres. We also evaluated variations of susceptibility rates by geographic region and infection type and assessed the prevalence of CPE-encoding genes among CREs.

Materials and methods

Bacterial isolates were collected via the SENTRY Antimicrobial Surveillance Program and sent to JMI Laboratories (North Liberty, IA, USA) for susceptibility testing [10]. Each participating centre was asked to collect a designated number of consecutive bacterial isolates per infection type, including bloodstream infection (BSI), pneumonia, skin and soft tissue infection (SSTI), urinary tract infection (UTI), and intra-abdominal infection (IAI). The number of isolates to be collected from each infection type was established by the study protocol, and the isolates were consecutively collected during a predetermined period of time, which was also specified by the study protocol and varied according to the type of infection. If a patient had more than one isolate, only the first isolate collected during the time period specified by the protocol was included in the study.

A total of 11,655 Enterobacterales isolates were collected consecutively in 2019 and 2020 from 38 medical centres located in Western Europe (W-EU; n = 8,784; 25 centres in 10 countries; Belgium, France, Germany, Ireland, Italy, Portugal, Spain, Sweden, Switzerland, and the UK) and the Eastern European and Mediterranean region (E-EU; n = 2,871; 13 centres in 10 countries; Belarus, Czech Republic, Greece, Hungary, Israel, Poland, Romania, Russia, Slovenia, and Turkey). Only isolates determined to be significant by local criteria as the reported probable cause of infection were included in this investigation. Species identification was confirmed by using standard biochemical tests and/or a MALDI Biotyper (Bruker Daltonics, Billerica, MA, USA), when necessary.

Carbapenem-resistant Enterobacterales (CRE) isolates were defined as displaying imipenem or meropenem MIC values at ≥ 4 mg/L. Imipenem was not applied to Proteus mirabilis or indole-positive Proteeae due to their intrinsically elevated MIC values. Isolates were categorised as MDR or XDR according to criteria defined in 2012 by the joint European and US Centers for Disease Control, which state MDR as nonsusceptible to ≥ 1 agent in ≥ 3 antimicrobial classes and XDR as susceptible to ≤ 2 classes [11]. The antimicrobial classes and drug representatives in this analysis included cephalosporins (ceftazidime, cefepime, and ceftriaxone), carbapenems (imipenem, meropenem, and doripenem), a broad-spectrum penicillin combined with a β-lactamase-inhibitor (piperacillin-tazobactam), fluoroquinolones (ciprofloxacin and levofloxacin), aminoglycosides (gentamicin, tobramycin, and amikacin), and a polymyxin (colistin).

Isolates were tested against aztreonam-avibactam and 12 comparator agents by the reference broth microdilution method specified by CLSI standards [12]. All tests were conducted in a central monitoring laboratory (JMI Laboratories). Aztreonam-avibactam was tested with avibactam at a fixed concentration of 4 mg/L. A tentative aztreonam-avibactam pharmacokinetic/pharmacodynamic (PK/PD) susceptible breakpoint of ≤ 8 mg/L was applied for comparison [1, 13]. EUCAST breakpoints were applied for the comparator agents where available [14]. The tigecycline susceptible breakpoint published by EUCAST for E. coli and C. koseri (≤ 0.5 mg/L) was applied to all Enterobacterales species for comparison. Concurrent quality control (QC) testing was performed to ensure proper test conditions and procedures. The QC strains tested included Escherichia coli ATCC 25,922 and ATCC 35,218; Klebsiella pneumoniae ATCC 700,603, ATCC BAA-1705, and ATCC BAA-2814; Pseudomonas aeruginosa ATCC 27,853; and Staphylococcus aureus ATCC 29,213.

All CRE isolates (n = 424) and the isolates with elevated aztreonam-avibactam MICs (> 8 mg/L; n = 6) were assessed for β-lactamase-encoding genes using next-generation sequencing (NGS), as previously described. [15] Furthermore, relative quantification of AmpC expression, the gene sequences encoding for OmpC and OmpF porins, and the penicillin-binding protein 3 (PBP3) were investigated in isolates with an elevated (> 8 mg/L) aztreonam-avibactam MIC [16].

Results

Aztreonam-avibactam activity was very consistent across the evaluated geographic regions and infection types. Aztreonam-avibactam inhibited 99.9% of Enterobacterales at ≤ 8 mg/L (n = 8,786; MIC50/90, ≤ 0.03/0.12 mg/L) and retained potent activity against CRE (n = 424; MIC50/90, 0.25/0.5 mg/L; 99.5% inhibited at ≤ 8 mg/L), MDR (n = 1,875; MIC50/90, 0.12/0.5 mg/L; 99.6% inhibited at ≤ 8 mg/L), and XDR isolates (n = 335; MIC50/90, 0.25/0.5 mg/L; 99.7% inhibited at ≤ 8 mg/L; Table 1).

Table 1 Antimicrobial activity of aztreonam-avibactam against the main organisms and organism groups
Full size table

A CPE encoding gene was identified in 360 of 424 (84.9%) CRE isolates (Table 1). The most common CPEs were blaKPC (154 isolates [36.3% of CRE], including blaKPC-2 [42] and blaKPC-3 [112]), followed by blaOXA-48 type (115 isolates [27.1% of CRE], including blaOXA-48 [92], blaOXA-48-like [1], blaOXA-232 [18], and blaOXA-181 [5]) and the MBLs (109 isolates [25.7% of CRE], including blaNDM-1 [81], blaNDM-5 [3], blaVIM-1 [22], blaVIM-19 [3], and blaIMP-1 [1]). Notably, 2 CPE-encoding genes were identified in 20 isolates, including one isolate with 2 blaOXA-48 type (blaOXA-48 and blaOXA-181) and one isolate with 2 MBL genes (blaNDM-1 and blaVIM-1; Table 1). Among MBL producers, 94 (86.2%) isolates were from E-EU and 15 (13.8%) isolates were from W-EU. All CPE-producer CRE isolates were inhibited at an aztreonam-avibactam ≤ 8 mg/L (MIC50/90, 0.25/0.5 mg/L; Table 1). Importantly, the highest aztreonam-avibactam MIC value among MBL-producing strains and among isolates producing 2 CPEs was only 0.5 mg/L (Table 1).

Aztreonam-avibactam was highly active against Enterobacterales isolates from W-EU (MIC50/90, ≤ 0.03/0.12 mg/L). Only 1 of 8784 isolates (0.01%) showed an aztreonam-avibactam MIC > 8 mg/L (MIC of 16 mg/L), an E. coli from Italy isolated from a patient with UTI (Table 2). The most active comparator agents against W-EU Enterobacterales isolates were meropenem (98.7% susceptible [S]), amikacin (97.7%S), and gentamicin (90.4%S; Table 2). Susceptibility varied slightly by infection type. The frequency of CRE, MDR, and XDR was the highest among BSI isolates while the CRE and MDR rates were lowest among UTI isolates and the XDR rate was lowest among SSTI isolates (Fig. 1).

Table 2 Antimicrobial activity of aztreonam-avibactam and comparator agents tested against Enterobacterales isolates from Western Europe (W-EU) and stratified by infection type
Full size table
Fig. 1
figure 1

Frequency of carbapenem-resistant Enterobacterales (CRE), multidrug-resistant (MDR), and extensively drug-resistant (XDR) isolates in Western Europe (W-EU) stratified by infection type

Full size image

Aztreonam-avibactam inhibited 99.9% of MDR isolates from W-EU at ≤ 8 mg/L; none of the comparator agents were active against > 90% of isolates (Table 2). The most active comparator agents against MDR isolates were meropenem, with susceptibility rates varying from 84.5% (BSI) to 93.8% (UTI; 88.5% overall), followed by amikacin (74.7–85.9%S; 80.7% overall) and colistin (69.5–80.8%S; 76.0% overall; Table 2). All CRE isolates from W-EU were inhibited at ≤ 8 mg/L of aztreonam-avibactam, and only colistin (90.6%S overall), amikacin (65.4%S), and tigecycline (52.8%S) were active against > 50% of W-EU CRE isolates (Table 2).

Aztreonam-avibactam was slightly (twofold) less active against isolates from E-EU (MIC50/90, 0.06/0.25 mg/L) compared to W-EU (MIC50/90, ≤ 0.03/0.12 mg/L). Only 5 of 2871 isolates (0.2%) from E-EU showed an aztreonam-avibactam MIC > 8 mg/L, 3 isolates from Poland (2 E. cloacae and 1 E. coli) and 2 from Turkey (2 E. coli; data not shown). The activities of the comparator agents were markedly lower against isolates from E-EU than W-EU (Tables 2 and 3). The most active comparator agents against E-EU isolates were meropenem (90.6%S), amikacin (88.2%S), and colistin (83.1%S; Table 3). Overall, CRE, MDR, and XDR rates were 10.3%, 32.0%, and 8.9% in E-EU and 1.4%, 10.9%, and 1.1% in W-EU, respectively (Figs. 1 and 2). The frequencies of CRE, MDR, and XDR in E-EU were highest among isolates from pneumonia and lowest among isolates from UTI (Fig. 2).

Table 3 Antimicrobial activity of aztreonam-avibactam and comparator agents tested against Enterobacterales isolates from Eastern Europe (E-EU) and stratified by infection type
Full size table
Fig. 2
figure 2

Frequency of carbapenem-resistant Enterobacterales (CRE), multidrug-resistant (MDR), and extensively drug-resistant (XDR) isolates in Eastern Europe (E-EU) stratified by infection type

Full size image

Percentages of E-EU MDR isolates inhibited at ≤ 8 mg/L of aztreonam-avibactam were 99.3% overall, and ranged from 99.1% (BSI and pneumonia) to 100.0% (IAI; Table 3). The most active comparator agents were colistin (69.0–83.2%S; 74.8% overall), meropenem (61.2–79.6%S; 70.8% overall), and amikacin (58.9–71.4%S; 64.3% overall; Table 3). Overall, 99.7% of CRE isolates from E-EU were inhibited at ≤ 8 mg/L of aztreonam-avibactam, including all isolates from SSTI, UTI, and IAI (Table 3). The most active comparator agents against E-EU CRE were colistin (74.0%S overall), gentamicin (40.1%S), and amikacin (38.7%S; Table 3).

Overall, only 6 of 11,655 (< 0.1%) Enterobacterales isolates tested showed an aztreonam-avibactam MIC > 8 mg/L: 4 E. coli and 2 E. cloacae. Results of the characterisation of these organisms are summarised in Table 4. Five organisms were from E-EU (Poland and Turkey) and 1 was from W-EU (Italy). A CPE-encoding gene was not detected in any of these isolates, except for a blaOXA-244 in E. coli 1,177,727. Four organisms were susceptible to meropenem (MIC, 0.06–0.5 mg/L) and ceftazidime-avibactam (MIC, 2–8 mg/L). Amino acid insertions and substitutions within PBP3 (YRIK) and a CMY-encoding gene were detected in all E. coli strains. In addition, these E. coli isolates carried multiple β-lactamase genes. Both E. cloacae overproduced AmpC (act-17 and act-24), carried ESBL genes, and had alterations in the porin sequence.

Table 4 Summary of the results on the characterisation of isolates exhibiting aztreonam-avibactam MIC results > 8 mg/L
Full size table

Discussion

Aztreonam-avibactam showed potent activity against a large collection of Enterobacterales isolates from W-EU and E-EU medical centres independent of infection type. Moreover, aztreonam-avibactam retained strong activity against CRE, including MBL producers, MDR, and XDR isolates. Our results corroborate those published by other investigators. Sonnevend et al. evaluated the activity of aztreonam-avibactam against 1192 CREs from 33 hospitals in 5 countries from the Arabian Peninsula. [17] Almost half (46.3%) of the isolates produced an MBL and 52.9% produced an OXA-48-like. Aztreonam-avibactam inhibited 95.5% of isolates at ≤ 4 mg/L and 46.7% of isolates were resistant to ceftazidime-avibactam. Notably, aztreonam-avibactam was active against 94.4% of ceftazidime-avibactam-resistant strains [17].

Resistance to aztreonam-avibactam (MIC, > 8 mg/L) was observed in only 6 isolates, 4 E. coli and 2 E. cloacae (Table 4). Decreased susceptibility to aztreonam-avibactam in E. coli has been reported by other investigators and seems to be caused by the association of PBP3 alterations and production of a CMY β-lactamase [16, 18,19,20]. Sadek et al. elegantly showed that the amino acid insertions YRIK and YRIN in the PBP3 protein are not sufficient to raise the aztreonam-avibactam MIC value to resistant levels (greater than 4 or 8 mg/L) and the presence of a blaCMY β-lactamase gene was associated with higher aztreonam-avibactam MIC values on isolates with those PBP3 alterations [20, 21]. All 4 aztreonam-avibactam-resistant E. coli isolates evaluated in this investigation showed an insertion of 4 amino acids (YRIK) in the PBP3 protein associated with a blaCMY gene. Plus, 2 isolates had a blaCMY-42, as reported by Sadek et al. [20, 21], whereas the other 2 isolates had blaCMY genes that differed from blaCMY-42 by only 1 amino acid, blaCMY-145 (N90T) and blaCMY-141 (I141L).

We did not identify β-lactamases known to be refractory to the avibactam inhibition, such PER or VEB, on the 2 aztreonam-avibactam-resistant E. cloacae isolates [19]. Moreover, the AmpC gene of one of the isolates (act-17) was cloned into an E. coli background and did not alter the aztreonam-avibactam MIC value of the recipient strain without this gene, which remained 0.12 mg/L [22]. Thus, we hypothesise that resistance to aztreonam-avibactam on these 2 E. cloacae strains was due to the association of AmpC hyperproduction and porin alterations.

Our results also showed that susceptibility to comparator agents varied between W-EU and E-EU and among infection types in each region. Resistance rates were generally higher in E-EU than W-EU. Notably, rates of CRE, MDR, and XDR were markedly higher among isolates from E-EU than W-EU (Figs. 1 and 2). These results clearly indicate a higher dissemination of ESBLs, CPEs, and other resistance mechanisms in E-EU compared to W-EU, corroborating the results from other large surveillance programmes. Results from previous SENTRY Program investigations as well as from those from other European surveillance programmes, such as the EARS-Net, have also shown a marked regional variation of antimicrobial resistance within Europe. Important antimicrobial resistance problems have been identified in many E-EU countries, such as Belarus, Greece, Poland, Russia, and Turkey [23,24,25,26,27,28].

Resistance rates also varied by infection type. In W-EU, resistance rates tended to be higher among isolates from BSI (13.0% MDR rate) and pneumonia (11.2% MDR rate) than other infection types, whereas in E-EU, resistance rates tended to be higher among isolates from pneumonia (39.2% MDR rate) and SSTI (33.4% MDR rates). Varying resistance rates by infection type have been reported by other investigators and could be related to several factors, including but not limited to underlying illness, duration of hospitalisation before acquiring the infection, or previous antibiotic exposure [29].

A few antimicrobial agents that are active against CRE have been licensed in the last few years, including ceftazidime-avibactam, meropenem-vaborbactam, imipenem-relebactam, and cefiderocol. Although the approval of these agents represented a remarkable progress in the treatment of infections caused by CRE, except for cefiderocol, these agents are not active against MBL-producing Enterobacterales [9, 30].

The results of this investigation revealed that 84.9% (360/424) of CRE isolates from this large European collection produced a CPE. Moreover, 30.3% (109/360) of CPE producers and 25.7% of CRE isolates (109/424) produced an MBL and are probably resistant to the β-lactam-β-lactamase inhibitors currently available, including ceftazidime-avibactam, meropenem-vaborbactam, and imipenem-relebactam.

Our results have some limitations. The fact that the criteria used to categorise a bacterial isolate as clinically significant were not defined in the study protocol and were based on local algorithms is a limitation since these criteria can vary among participating medical centres. Also, we could not differentiate between subsets of infection types that may present different susceptibility patterns, such as catheter-related versus non-catheter-related BSI or surgical versus non-surgical SSTI. Finally, this study had a restricted number of medical centres in some countries. These limitations should be considered when interpreting the results and conclusions.

In conclusion, resistance to aztreonam-avibactam was extremely rare among a large collection of Enterobacterales from European medical centres. The results of this large, international investigation support the clinical development of aztreonam-avibactam for treatment of Enterobacterales infections, including those infections caused by MBL-producing strains.