In vitro Activity of Cefiderocol and Comparators against Carbapenem-Resistant Gram-Negative Pathogens from France and Belgium

Infections with carbapenem-resistant (CR) Gram-negative (GN) pathogens have increased in many countries worldwide, leaving only few therapeutic options. Cefiderocol (CFDC) is approved in Europe for the treatment of aerobic GN infections in adults with limited treatment options. This study evaluated the in vitro activity of cefiderocol and comparators against multidrug-resistant (MDR) bacteria including meropenem-resistant (MR) or pandrug-resistant (PR) GN clinical isolates from France and Belgium. The minimum inhibitory concentrations (MICs) of CFDC were determined by broth microdilution, using iron-depleted cation-adjusted Mueller–Hinton broth, and were compared to those of 10 last-line antibiotics. The MICs were interpreted according to EUCAST and CLSI breakpoints, and in the absence of species-specific breakpoints, non-species-related pharmacokinetic/pharmacodynamic breakpoints were used. Among the 476 isolates tested, 322 were carbapenemase producers (CP), 58 non-CP-CRs, 52 intrinsically CR, 41 expanded-spectrum cephalosporin resistant and 5 were multi-susceptible. Susceptibility to CFDC was high using EUCAST breakpoints 81%, 99% and 84%, and was even higher using CLSI breakpoints to 93%, 100% and 88% for Enterobacterales, Pseudomonas aeruginosa and Acinetobacter baumannii, respectively. Susceptibility to cefiderocol using non-species-related breakpoints for Stenotrophomonas maltophilia, Achromobacter xylosoxydans and Burkholderia cepacia, was 100%, 100% and 92.3%, respectively. The susceptibility rates were lower with the NDM producers, with values of 48% and 30% using EUCAST breakpoints and 81% and 50% using CLSI breakpoints for Enterobacterales and Acinetobacter spp, respectively. CFDC demonstrated high in vitro susceptibility rates against a wide range of MDR GN pathogens, including MR and PR isolates.

Here, we report the antimicrobial activity of CFDC and comparators (aztreonam, amikacin, cefepime, ceftazidime, ceftazidime-avibactam, ceftolozane-tazobactam, ciprofloxacin, meropenem, colistin and tigecycline) against a panel of 476 mostly MDR GNB collected from hospitals in France and Belgium between 2012 and 2019, thus before any clinical use of CFDC (Table 1).

Activity of Cefiderocol
The in vitro activity of CFDC and comparators was assessed in 476 Gram-negative isolates collected from two National Reference Centers (NRC) for AMR located in France and Belgium (Table 1). The 222 (46.6%) Enterobacterales isolates were from the French NRC for CREs and the remaining 254 (53.4%) bacteria came equally from the two NRCs. These isolates were MDR and of reduced susceptibility/resistant to carbapenems.
A total of 92% of the KPC producers were susceptible to CFDC, with an MIC 50/90 of 1/4 mg/L, results that are similar to those of ceftazidime/avibactam (94% susceptibility, and an MIC 50/90 of 2/8). The only other competitive comparators were amikacin (76%), tigecycline (76%) (0.5/1) and colistin (62%) (0.5/>8). For all the other antibiotics, the MIC 50/90 values were superior or equal to the upper limit of the concentration range used in the MIC testing.
Among the 120 P. aeruginosa isolates tested, only one isolate exhibited an MIC value of CFDC of 4 mg/L, which is considered resistant by EUCAST, but still susceptible by CLSI. Noteworthy, in P. aeruginosa, CFDC was active against all MBL producers, while all comparators were below 20% except for colistin (97%). Thirteen A. baumannii (15.9%) had MICs > 2 mg/L, among which seven were NDM producers, three were ESBL producers (two PER and one VEB) and three were OXA-23 producers.

Discussion
In the French SIDERO-WT study, CFDC displayed excellent in vitro activity, but since only very few meropenem-resistant Enterobacterales were included, it was not possible to assess CFDC activity versus other antimicrobials as comparators for these isolates [29]. In the present evaluation of 476 GNB isolates, of which 472 were MDR, from France and Belgium (of which 66% were meropenem R (MIC > 8 mg/L)), CFDC demonstrated substantial in vitro activity. These isolates were from hospitals in France and Belgium and were mainly isolated in 2018 and 2019, thus before any clinical use of CFDC. Notably, CFDC demonstrated substantial activity against all isolates of P. aeruginosa, most A. baumannii and intrinsically meropenem-resistant GN non-fermenters such as S. maltophilia and A. xylosoxidans, where all other comparators demonstrated much lower susceptibility rates. Overall, more isolates were susceptible to CFDC than to a key subset of other currently available antimicrobial agents including the β-lactam/β-lactamase inhibitor combinations ceftazidime-avibactam and ceftolozane-tazobactam, and colistin, which is often considered a last resort molecule for MDR GNB infections.
Carbapenem resistance among GNB in France is steadily rising and poses a substantial threat to patients and healthcare systems, often leading to greater rates of mortality, morbidity and increased burden on hospitals [8,9,11]. In France and Belgium, OXA-48 remains the most common carbapenemase reported nationally among CPEs [8,9,11]. The OXA-48 variants, OXA-181 and the difficult-to-detect OXA-244 are increasingly isolated among CPEs, including ESBL producers. There is now increasing concern over the emergence of OXA-48-mediated resistance to new antibiotic regimens such as ceftazidime-avibactam [34]. Therefore, there is a continued need for new antibiotics and antibiotic regimens with activity against OXA-48-like producers, among others. Despite the dominance of OXA-48 in France and Belgium, over recent years there has been a notable shift in resistance mechanisms, with an increase in MBL producers such as NDM and VIM [8,9,11] and the first isolation of IMP-producing Enterobacterales. This shift to MBL-mediated resistance is of concern as new β-lactam/β-lactamase inhibitor combination therapies, including ceftazidime-avibactam, ceftolozane-tazobactam and meropenem-vaborbactam, are known to lack efficacy against MBLs. As such, these agents cannot be proposed for empirical treatment against infections that are suspected to involve MBL-producing GNB [35,36]. Previous reports from the SIDERO-WT study have shown potent in vitro activity of CFDC against carbapenemaseproducing isolates including MBL producers [27,29]. In these studies, only a few MBL producers were included. Here, in our study, 54 MBLs, among which were 21 NDM producers, were studied. Only 50% of the NDM-producing Enterobacterales were susceptible, which was yet much higher than the comparators, except for colistin and tigecycline. In general, MICs for cefiderocol with Enterobacterales-producing MBLs are close to the breakpoints with 28% (n = 15) with an MIC = 1 mg/L, 20% (n = 11) with an MIC = 2 mg/L, and 20% (n = 11) with an MIC = 4 mg/L. A possible explanation for these higher MICs as compared to other carbapenemases [37] is likely due to the fact that MBLs and especially NDM have stronger hydrolytic activity against expanded-spectrum cephalosporins, including CFDC, as suggested by the addition the MBL inhibitor, dipicolinic acid, that reduced the MICs of CFDC against previously non-susceptible Enterobacterales isolates [38]. Based on resistance reports, the increased copy number of NDM may increase CFDC MIC values in the absence of CirA mutations, which is the iron transporter involved in CFDC uptake. However, when NDM overexpression is associated with mutations of the cirA gene, a loss of fitness was observed in these isolates. Of note, the combination of mutations in the iron transport genes and the expression of the NDM enzyme was found in CREs in China, way before CFDC was used in clinical practice [39]; thus, it has been suggested that resistance to CFDC may be the consequence of previous antibiotic treatments, to cancer therapies or to so far unknown mechanisms of selection [40][41][42].
Furthermore, in instances where resistance is not due to carbapenemase production, CFDC has demonstrated in vitro activity against isolates with AmpC, ESBLs, porin mutations and efflux pump upregulation.
Although the prevalence of infections caused by non-fermenting GNB currently remains relatively low in France, there is growing concern regarding the high propensity of these isolates to develop resistance, and the resulting depletion of available effective treatment options [13,43]. In this study, the CFDC activity exceeded that of all the tested comparators except colistin against meropenem-resistant isolates of S. maltophilia, P. aeruginosa and A. baumannii.
These findings are in line with previous reports in other countries, which demonstrated the potent in vitro activity of CFDC against MDR Enterobacterales, MDR A. baumannii, MDR P. aeruginosa and S. maltophilia [37,44]. Additionally, novel agents such as ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam have recently been approved against antibiotic-resistant GNB as they are effective against Enterobacterales-producing KPC but have limited or no efficacy against CR A. baumannii [35,43].
Overall, there are very few antimicrobial agents available to clinicians to treat patients infected with CR GNB, and the agents that are available are often associated with considerable toxicities and increasing resistance. Colistin is effective against a wide range of CR-GNB, and in this study, colistin was the only agent with comparable activity to CFDC against non-fermenters collected from patients with nosocomial pneumonia or bloodstream infections (BSI). However, the usage of colistin is associated with a potential risk of nephroand neurotoxicity [45] and several species of Enterobacterales have demonstrated intrinsic colistin resistance. Additionally, in this study, fewer meropenem-resistant S. maltophilia isolates were susceptible to colistin than to CFDC. The in vivo results confirmed the excellent behavior of CFDC for the treatment of MDR GNB in bloodstream infections [46][47][48][49][50][51]. CFDC has also shown to be a promising new treatment option for patients with bone and joint infections due to CR A. baumannii and appears to be well tolerated for prolonged durations [49,50].

Bacteria
The tested isolates (Table 1) were from the French and Belgium National Reference Centers for antibiotic resistances among GNB and comprised (i) 222 isolates of Enterobacterales, selected to represent diverse carbapenemase producers and isolates with carbapenem resistance via combinations of porin loss with AmpC or ESBL activity; (ii) 120 isolates of P. aeruginosa, selected to represent producers of MBLs and GES carbapenemases, along with isolates that produced ESBLs and were carbapenem-resistant via porine OprD loss; (iii) 82 MDR isolates of A. baumannii expressing various carbapenemases, including NDM and/or various OXA carbapenemases; and (iv) 52 GNB naturally resistant to carbapenems: 25 S. maltophilia, 13 B. cepacia, 12 A. xylosoxidans and 2 Elizabethkingia sp.
These isolates were selected by both NRCs to represent the French and Belgium epidemiology of carbapenem-resistant GNB and challenging isolates expressing rare carbapenemases. As CFDC had not previously been tested, these isolates were chosen based on their carbapenem/expanded spectrum susceptibility profiles and their enzymatic content. Almost all the isolates tested were submitted for an investigation of MDR/XDR resistance phenotypes by hospital laboratories in France and Belgium between 2012 and 2019, thus before any clinical use of CFDC. Carbapenemases and ESBL enzymes were identified by PCR of their encoding genes or by whole-genome sequencing (WGS). Carbapenem resistance due to porin loss combined with ESBL or AmpC activity was inferred from previous susceptibility results and the absence of carbapenemase, as confirmed by PCR or WGS. Species identification was by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectroscopy.

Antimicrobial Susceptibility Testing
The MICs were determined using frozen 96-well broth microdilution panels with a pre-loaded antibiotic growth medium supplied by International Health Management Associates, Inc. (IHMA; Schaumburg, IL, USA). CFDC was tested in iron-depleted cationadjusted Mueller-Hinton broth (ID-CAMHB), as recently approved by the CLSI ( [32]; http://clsi.org/standards/micro/microbiology-files/, accessed on 1 September 2022), whereas the comparators were tested in cation-adjusted Mueller-Hinton broth (CAMHB). The strains were grown overnight on a non-selective agar media. Two to three colonies were resuspended in 3 mL sterile 0.85% NaCl in order to obtain a 0.5 McFarland suspension. One milliliter of this suspension was further diluted in 29 mL of sterile water, of which 10 µL were then added to each well, and the plates were subsequently incubated for 16-20 h at 35 • C, as recommended by the manufacturer and EUCAST guidelines [31]. Quality control testing was performed on each day of testing using E. coli ATCC 25922, K. pneumoniae ATCC 700603 and P. aeruginosa ATCC 27853 to ensure the stability of the panels and the validity of the test methods. The comparator antibiotics were for Enterobacterales, Pseudomonas spp., Acinetobacter spp., B. cepacia, A. xylosoxydans and Elizabethkingia spp. meropenem, ceftazidime, ceftazidime-avibactam (4 µg/mL), cefepime, ceftolozane-tazobactam (4 µg/mL), aztreonam, colistin, amikacin, ciprofloxacin and tigecycline, all sourced by IHMA. For S. maltophilia, cotrimoxazole, levofloxacin and minocycline were tested instead of cefepime, aztreonam and ciprofloxacin.
The MIC results of CFDC were interpreted using EUCAST breakpoint [31] values of S ≤ 2 mg/L and R > 2 mg/L for Enterobacterales, P. aeruginosa, Acinetobacter spp and S. maltophilia, and for the other tested bacteria, non-species-related PK/PD values (≤2 mg/L ) were used; the Investigational CLSI MIC breakpoints for the same bacteria were used with values of S ≤ 4 mg/L and R ≥ 16 mg/L, which correspond to those when CFDC was in trial. The MICs of the comparator antibiotics were interpreted using EUCAST guidelines where available, the exceptions being ceftazidime and cefepime for Acinetobacter spp., for which only the CLSI breakpoints are available [32,33].

Ethics
Ethics approval was not required as all the bacterial isolates were from the French or Belgium NRC for antibiotic resistances and thus were anonymized and unrelated to the patients.

Conclusions
The increasing incidence and diversification of carbapenem resistance among GNB is of growing concern in France and in Belgium, as a shift toward more difficult-totreat pathogens is putting pressure on the already limited available treatment options. CFDC demonstrates substantial and broad in vitro activity against a wide range of MDR pathogens, and even XDR GNB. The findings from this study are in line with those from previous reports and suggest that CFDC may offer an invaluable treatment option in the fight against antimicrobial-resistant GNB, particularly for carbapenem-resistant non-fermenters and MBL producers, especially Acinetobacter baumannii, for which there are currently few approved effective therapies. Colistin was the only other agent with similar activity as CFDC against meropenem-resistant GNB. It should be emphasized that CFDC displays much more favorable pharmacokinetic parameters (tissue diffusion and use in renal impairment) than colistin and tigecycline, which will be an important factor for choosing an adequate therapy for infections due to multidrug infections.
In addition to aztreonam, CFDC is the other beta-lactam with activity against MBLproducing CREs. Our results, along with other in vitro and surveillance studies, showed that CFDC MIC values are higher against NDM-producing isolates than VIM-producing isolates. Nevertheless, clinical studies demonstrated that NDM-producing CRE infections with CFDC MICs of 4 µg/mL, which corresponds to the CLSI susceptibility breakpoint, could be successfully treated [48]. The recent IDSA guidance and ESCMID guidelines provide recommendations on when and how to use the new antimicrobial agents, especially to prevent irrational use and the emergence of resistance [52,53]. Neither of them recommends a second agent to be used with the new antibiotics for the treatment of CRE infections. Even though resistance for each of the new agents has been described, great susceptibility rates are described globally, with some regional variations. Overall susceptibility rates are reduced for ceftazidime-avibactam, meropenem-vaborbactam and imipenem-relebactam in regions where MBLs are prevalent, and CFDC MICs are higher where NDM-producing CREs are more prevalent. This underlines the need for rapid diagnostic tests for resistance mechanisms that will improve the surveillance and diagnosis of CRE and, hence, the selection of the most appropriate antibiotic agent [54,55].  Conflicts of Interest: T.N. reports non-financial support from Pfizer, personal fees and non-financial support from Shionogi, outside the submitted work. C.L. is a fulltime employee of Shionogi B.V., London, UK. All other authors have nothing to disclose. The funders had no role in the collection, analyses or interpretation of the data; in the writing of the manuscript; or in the decision to publish the results.