In Vitro Susceptibility of Gram-Negative Pathogens to Cefiderocol in Five Consecutive Annual Multinational SIDERO-WT Surveillance Studies, 2014 to 2019

The prevalence of carbapenem-resistant, multidrug-resistant (MDR), and difficult-to-treat Gram-negative bacilli is increasing worldwide, and therapeutic options for infected patients are often limited (1–3). The World Health Organization (WHO) has classified carbapenem-resistant Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii as pathogens of the highest (critical) priority for development of new antibacterial agents (2). Cefiderocol, a parenteral siderophore cephalosporin, was approved by the United States Food and Drug Administration (FDA) in November 2019 for the treatment of adults with complicated urinary tract infections, including pyelonephritis, caused by susceptible Gram-negative bacilli (Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Enterobacter cloacae complex, and P. aeruginosa) when limited or no other treatment options exist (4). In April 2020, the European Medicines Agency (EMA) licensed cefiderocol for the treatment of infections due to aerobic Gram-negative organisms in adults with limited treatment options (5). In September 2020, the FDA approved cefiderocol for a new indication, the treatment of hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia caused by Enterobacterales (E. coli, K. pneumoniae, E. cloacae complex, and Serratia marcescens), P. aeruginosa, and Acinetobacter baumannii complex. Clinical development of cefiderocol continues for the treatment of serious infections attributable to resistant Gram-negative bacilli, including infections caused by carbapenem-resistant Gram-negative bacilli (6).

Cefiderocol possesses a unique mechanism of bacterial cell entry, making it an important addition to the antimicrobial armamentarium. The optimized chloro-catechol moiety within the C-3 side chain of cefiderocol facilitates formation of chelated complexes with ferric iron and expedites its transport across the outer membrane of Gram-negative bacilli using constitutive iron transport systems (7). Following its delivery to the periplasmic space, cefiderocol binds primarily to penicillin binding protein 3 (PBP 3), similarly to other cephalosporins, and impedes peptidoglycan synthesis (7). Cefiderocol has been shown to be stable to hydrolysis by most clinically important β-lactamases, including both serine β-lactamases of Ambler classes A (e.g., KPC and, extended-spectrum lactamase [ESBL; e.g., CTX type, SHV type, and TEM type]), C (i.e., AmpC), and D (e.g., OXA) carbapenemases and metallo-β-lactamases of Ambler class B (e.g., IMP, NDM, and VIM) and to be minimally affected by porin deletions and efflux-mediated resistance mechanisms (2- to 4-fold increases in cefiderocol MIC) (7–14).

Using current standardized reference testing methods and reliable, predictable, evidence-driven MIC and disk diffusion zone size interpretative criteria to determine in vitro activities for recently approved and investigational agents is critical to establishing and supporting treatment decisions and expanding the role of these agents in patient care, particularly for patients where unmet medical need exists (2). Investigational MIC and disk diffusion zone diameter interpretative criteria for cefiderocol were published by the Clinical and Laboratory Standards Institute (CLSI) in 2019 based on in vitro activity and preclinical in vivo pharmacokinetic/pharmacodynamics data prior to FDA approval of cefiderocol (15, 16). In February 2021, CLSI approved MIC clinical breakpoints for Enterobacterales, P. aeruginosa, and Acinetobacter species of ≤4 μg/mL (susceptible), 8 μg/mL (intermediate), and ≥16 μg/mL (resistant) and for Stenotrophomonas maltophilia of ≤1 μg/mL (susceptible) and >1 μg/mL (nonsusceptible) (17). The updated MIC clinical breakpoints for cefiderocol will be published in early 2022 with the release of the 32nd edition of the CLSI M100 document. Clinical breakpoints for cefiderocol are also available from the United States Food and Drug Administration (FDA) (18) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) (19) (see Table S1 in the supplemental material) but were not used for MIC interpretation in this article.

The intent of the current study was to evaluate the in vitro susceptibility to cefiderocol of Gram-negative pathogens (Enterobacterales, P. aeruginosa, A. baumannii complex, S. maltophilia, and Burkholderia cepacia complex) collected over five consecutive annual SIDERO-WT surveillance studies (from November 2014 to December 2019) conducted in North America and Europe using the recently approved (February 2021) CLSI MIC clinical breakpoints (17). In addition, we performed cefiderocol susceptibility subset analysis that included isolates with meropenem-, ceftazidime-avibactam-, and ceftolozane-tazobactam-nonsusceptible phenotypes, as it is in patients infected with these nonsusceptible isolates where cefiderocol use most directly addresses an unmet medical need, and reviewed the cefiderocol percent susceptible rates and isolate MIC distributions by year to identify trends in cefiderocol in vitro activity over time.

The minimal inhibitory concentrations of cefiderocol that inhibited 50% (MIC₅₀) and 90% (MIC₉₀) of the 31,896 isolates of Enterobacterales tested from North America and Europe from 2014 to 2019 were 0.12 and 1 μg/mL, respectively (Table 1). Cefiderocol inhibited 99.8% of all isolates of Enterobacterales at a MIC of ≤4 μg/mL. The cefiderocol MIC₅₀ and MIC₉₀ were 1 and 4 μg/mL, respectively, for the subset of 1,021 meropenem-nonsusceptible (MIC, ≥2 μg/mL) isolates of Enterobacterales; 96.7% of meropenem-nonsusceptible isolates were susceptible to cefiderocol. Cefiderocol demonstrated a higher percent susceptible rate against meropenem-nonsusceptible isolates (≥20% higher) than ceftazidime-avibactam (77.0%), cefepime (8.7%), ceftolozane-tazobactam (7.8%), and ciprofloxacin (7.8%). A total of 91.6% of 263 isolates of ceftazidime-avibactam-nonsusceptible (MIC, ≥16 μg/mL) Enterobacterales and 97.7% of 2,658 isolates of ceftolozane-tazobactam-nonsusceptible (MIC, ≥4 μg/mL) Enterobacterales were susceptible to cefiderocol. In comparison, only 3.8% of ceftazidime-avibactam-nonsusceptible Enterobacterales isolates were susceptible to ceftolozane-tazobactam and 90.5% of ceftolozane-tazobactam-nonsusceptible Enterobacterales isolates were susceptible to ceftazidime-avibactam. MIC₉₀ values for colistin (excluding isolates with intrinsic resistance—Proteus spp., Providencia spp., Morganella morganii, and S. marcescens) and ciprofloxacin were 1 and >8 μg/mL, respectively, for all isolates of Enterobacterales tested.

Cefiderocol MICs for meropenem-nonsusceptible (Fig. 1), ceftazidime-avibactam-nonsusceptible (Fig. 2), and ceftolozane-tazobactam-nonsusceptible (Fig. 3) isolates of Enterobacterales (both North American and European isolates combined) demonstrated a rightward shift (of 1 to 3 doubling dilutions) to higher cefiderocol MICs compared to each respective antimicrobial-susceptible subset; however, as mentioned earlier, most meropenem (96.7%)-, ceftazidime-avibactam (91.6%)-, and ceftolozane-tazobactam (97.7%)-nonsusceptible isolates remained susceptible to cefiderocol, with MICs of ≤4 μg/mL.

The cefiderocol MIC₅₀ and MIC₉₀ were 0.12 and 0.5 μg/mL for 7,700 isolates of P. aeruginosa collected in North America and Europe from 2014 to 2019 (Table 1). Cefiderocol inhibited 99.9% of isolates at ≤4 μg/mL. Ceftazidime-avibactam (93.8% susceptible) and ceftolozane-tazobactam (94.0% susceptible) were less active than cefiderocol against all isolates of P. aeruginosa tested. The MIC₅₀ and MIC₉₀ values for cefiderocol against the subset of 1,759 isolates of meropenem-nonsusceptible (MIC, ≥4 μg/mL) P. aeruginosa were 0.25 and 1 μg/mL, respectively, and 99.9% of meropenem-nonsusceptible isolates were susceptible to cefiderocol. Ceftazidime-avibactam, ceftolozane-tazobactam, and cefepime were all tested with MIC₉₀ values of 32 or >64 μg/mL against isolates of meropenem-nonsusceptible P. aeruginosa and exhibited percent susceptible rates of 76.1% (ceftolozane-tazobactam), 75.0% (ceftazidime-avibactam), and 49.0% (cefepime). MIC₉₀ values for ciprofloxacin (31.2% susceptible) and colistin were >8 μg/mL and 1 μg/mL, respectively, for meropenem-nonsusceptible P. aeruginosa. A total of 100% of 477 isolates of ceftazidime-avibactam-nonsusceptible (MIC, ≥16 μg/mL) and 99.8% of 463 isolates of ceftolozane-tazobactam-nonsusceptible (MIC, ≥8 μg/mL) P. aeruginosa, respectively, were susceptible to cefiderocol. In comparison, only 24.3% of ceftazidime-avibactam-nonsusceptible P. aeruginosa isolates were susceptible to ceftolozane-tazobactam, and only 22.0% of ceftolozane-tazobactam-nonsusceptible P. aeruginosa isolates were susceptible to ceftazidime-avibactam.

Cefiderocol MICs for meropenem-nonsusceptible (Fig. 1), ceftazidime-avibactam-nonsusceptible (Fig. 2), and ceftolozane-tazobactam-nonsusceptible (Fig. 3) isolates of P. aeruginosa (both North American and European isolates combined) demonstrated a rightward shift (of 1 doubling dilution) to higher cefiderocol MICs compared to each respective antimicrobial-susceptible subset; however, almost every nonsusceptible isolate (99.8 to 100%) remained susceptible to cefiderocol, with a MIC of ≤4 μg/mL.

The MIC₅₀ and MIC₉₀ of cefiderocol for isolates of A. baumannii complex from both North America and Europe were 0.12 and 1 μg/mL; 96.0% of isolates demonstrated cefiderocol MICs of ≤4 μg/mL (Table 1). Cefiderocol MIC distributions for meropenem-nonsusceptible isolates demonstrated approximately a 1-doubling-dilution rightward shift to higher MICs relative to the meropenem-susceptible isolate subset (Fig. 1); 94.2% of 2,810 isolates of meropenem-nonsusceptible A. baumannii complex remained cefiderocol susceptible, with a MIC₉₀ value of 2 μg/mL. Ceftazidime-avibactam, ceftolozane-tazobactam, cefepime, and ciprofloxacin were inactive against A. baumannii complex. The MIC₉₀ for colistin against meropenem-nonsusceptible A. baumannii complex was >8 μg/mL, and 12.8% of isolates were colistin resistant. Cefiderocol also inhibited 98.6% of S. maltophilia isolates at ≤1 μg/mL (Table 1). B. cepacia complex isolates tested with cefiderocol MIC₅₀ and MIC₉₀ values of ≤0.03 and 0.5 μg/mL within 1 doubling-dilution of the MIC₅₀ (≤0.03 μg/mL) and MIC₉₀ values (1 μg/mL) for the meropenem-nonsusceptible subset of isolates (Table 1).

Annual cefiderocol percent susceptible rates for isolates of Enterobacterales from North America (99.6 to 100% susceptible) and Europe (99.3 to 99.9% susceptible) varied over very narrow ranges (0.4 to 0.6%) (Table 2). Even less variation (0.1 to 0.2%) in annual cefiderocol percent susceptible rates was observed for P. aeruginosa. Annual cefiderocol percent susceptible rates for isolates of P. aeruginosa from North America ranged from 99.8% to 100%, and those for isolates from Europe ranged from 99.9% to 100%. Annual percent susceptible rates for A. baumannii demonstrated sporadic, nondirectional differences. The annual cefiderocol percent susceptible rate range was narrower for isolates of A. baumannii from North America (97.5 to 100%) than for isolates from Europe (90.4 to 97.5%). In total, there were 171 isolates of A. baumannii with cefiderocol MICs of ≥8 μg/mL (nonsusceptible) collected in Europe from 2014 to 2019. Of these isolates, 74.3% (127/171) were from one country (Russia); 127/437 (29.1%) of isolates from Russia were cefiderocol nonsusceptible, with annual rates of 28.2% (11/39) in 2014, 41.2% (7/17) in 2015, 24.1% (19/79) in 2016, 42.7% (50/117) in 2017, 31.9% (36/113) in 2018, and 5.6% (4/72) in 2019. Other European countries contributing >10 isolates over the study period submitted isolates with cefiderocol-nonsusceptible MICs at rates ranging from zero (no cefiderocol-nonsusceptible isolates) to 7.3% (8/109 isolates [United Kingdom]). Annual cefiderocol MIC distributions for Enterobacterales, P. aeruginosa, A. baumannii complex, S. maltophilia, and B. cepacia complex are provided in Tables S2 to S6 in the supplemental material.

Annual percent susceptible rates for ceftazidime-avibactam for isolates of Enterobacterales from North America (99.7 to 100%) and Europe (98.2 to 98.8%) were similar (<2% annual variation), while annual percent susceptible rates for ceftolozane-tazobactam were higher in isolates from North America (93.8 to 94.9%) than in those from Europe (87.3 to 90.6%) (Table 2). Annual percent susceptible rates for isolates of P. aeruginosa from North America were higher for both ceftazidime-avibactam (96.0 to 99.6%) and ceftolozane-tazobactam (96.7 to 99.6%) than for isolates from Europe (ceftazidime-avibactam, 90.3 to 93.1%; ceftolozane-tazobactam, 90.2 to 91.5%).

Isolates of Enterobacterales, P. aeruginosa, A. baumannii complex, S. maltophilia, and B. cepacia complex collected in 2019 were also tested against meropenem-vaborbactam and imipenem-relebactam (see Table S7 in the supplemental material). Meropenem-vaborbactam demonstrated in vitro activity similar to that of ceftazidime-avibactam against Enterobacterales (98.9% of isolates susceptible); <70% of meropenem-nonsusceptible Enterobacterales isolates were susceptible to meropenem-vaborbactam and imipenem-relebactam, compared to 93.2% susceptible for cefiderocol. Imipenem-relebactam was less active (83.9% susceptible) than ceftazidime-avibactam against P. aeruginosa, compared to 99.9% susceptible for cefiderocol. Meropenem-vaborbactam and imipenem-relebactam were largely inactive in vitro against clinical isolates of A. baumannii complex (MIC₉₀, >16 μg/mL) and S. maltophilia (MIC₉₀, >16 μg/mL).

Data in the current study clearly demonstrate that the large majority of isolates of Enterobacterales (99.8%), P. aeruginosa (99.9%), A. baumannii complex (96.0%), and S. maltophilia (98.6%) collected across North America and Europe from 2104 to 2019 were susceptible to cefiderocol. Data in the current study confirm and expand upon data presented in earlier studies. Cefiderocol was previously reported to demonstrate potent in vitro activity against key Gram-negative pathogens (Enterobacterales, P. aeruginosa, Acinetobacter, Stenotrophomonas, and Burkholderia) but only limited activity against Gram-positive and anaerobic bacteria (7, 10, 15). International and regional surveillance studies (10, 15, 20–23) and resistant isolate collection profiling studies (8, 9, 11–14, 24) have reported ≥99% of Enterobacterales, P. aeruginosa, and S. maltophilia isolates and ≥96% of A. baumannii complex isolates have cefiderocol MICs of ≤4 μg/mL (10, 15, 20–23). Cefiderocol MICs were also ≤4 μg/mL for most carbapenem-resistant Enterobacterales (≥95% of isolates), P. aeruginosa (≥97%), and A. baumannii complex (≥91%) isolates, as well as MDR Enterobacterales (≥97%), P. aeruginosa (≥97%), and A. baumannii complex (≥90%) isolates (8–10, 14, 15, 20–24). Cefiderocol has dependably shown in vitro potency superior to those of ceftazidime-avibactam, ceftolozane-tazobactam, cefepime, ciprofloxacin, and colistin against clinical isolates of meropenem-resistant Enterobacterales, P. aeruginosa, and A. baumannii complex and to inhibit almost all isolates of Enterobacterales (>98%) and P. aeruginosa (>99%), with ceftazidime-avibactam-, ceftolozane-tazobactam-, cefepime-, ciprofloxacin- and colistin-resistant phenotypes at MICs of ≤4 μg/mL (8–10, 15, 20, 21, 24). Importantly, there was no appreciable cross-resistance between cefiderocol and ceftazidime-avibactam, ceftolozane-tazobactam, meropenem, or cefepime for Enterobacterales or P. aeruginosa, even though all are β-lactams. Most isolates resistant to newer β-lactam/β-lactamase inhibitor combinations remain susceptible to cefiderocol. In the current study, 91.6% of isolates of ceftazidime-avibactam-nonsusceptible Enterobacterales and 97.7% of isolates of ceftolozane-tazobactam-nonsusceptible Enterobacterales were susceptible to cefiderocol, as were 100% of isolates of ceftazidime-avibactam-nonsusceptible P. aeruginosa and 99.8% of isolates of ceftolozane-tazobactam-nonsusceptible P. aeruginosa. The current study also found that 93.9% of all isolates of B. cepacia complex had cefiderocol MICs of ≤1 μg/mL, and 95.5% of isolates had MICs of ≤4 μg/mL, similar to previous reports (10, 15, 20). We also confirmed that ceftazidime-avibactam and ceftolozane-tazobactam are largely inactive in vitro against clinical isolates of A. baumannii complex (MIC₉₀, >64 μg/mL), S. maltophilia (MIC₉₀, 64 to >64 μg/mL), and B. cepacia complex (MIC₉₀, 8 to 32 μg/mL).

Single, specific mechanisms conferring resistance to cefiderocol in Enterobacterales, P. aeruginosa, and A. baumannii have not been identified, although the addition of avibactam, a β-lactamase inhibitor, to cefiderocol has been shown to lower the MICs for some cefiderocol-resistant isolates, primarily A. baumannii possessing various ESBLs. (15, 25, 26). In addition, in some isolates of Gram-negative bacilli with cefiderocol MICs ranging from 2 to 256 μg/mL, the addition of a β-lactamase inhibitor (e.g., clavulanic acid, avibactam, or dipicolinic acid) was shown to lower cefiderocol MICs (4). Cross-resistance between cefiderocol and other antibacterial classes has not been identified; generally, isolates of Gram-negative bacilli resistant to other antibacterial agents are reliably susceptible to cefiderocol (10, 15, 20). The frequency of resistance development in Gram-negative bacteria, including carbapenemase producers exposed to cefiderocol at 10 times the MIC, ranged from 10⁻⁶ to 10⁻⁸ (7, 27). Mutations in the upstream region of pvdS and fecI in P. aeruginosa, which could affect the expression of ferric siderophore uptake-related genes, were reported to increase cefiderocol MICs by 32-fold (7). Overproduction of AmpC, modifications of PBPs, and loss of the TonB energy-transducing protein or the siderophore receptors CirA and Fiu (Enterobacterales) or PiuA (not PirA) (P. aeruginosa) can also elevate cefiderocol MICs (4, 8). Cefiderocol does not induce AmpC β-lactamase production in P. aeruginosa and E. cloacae (28).

Even though resistance to cefiderocol has not been observed to be consistently mediated by the presence of specific carbapenemases (14), higher cefiderocol MICs among NDM-positive and PER-positive Gram-negative bacilli than among isolates producing other carbapenemases has been observed (9, 11, 14, 15). However, many isolates of NDM-producing Enterobacterales demonstrated cefiderocol MICs of ≤4 μg/mL (9, 11, 14, 15), and infections with NDM-producing Enterobacterales have been treated effectively with cefiderocol, as observed in clinical trials (29).

Different MIC breakpoints for cefiderocol have been published (16–19). Determining the in vitro susceptibility of clinical isolates to cefiderocol would benefit from the application of a uniform set of MIC and/or disk diffusion breakpoints. With the recent CLSI approval of clinical breakpoints for cefiderocol (17), CLSI and FDA MIC breakpoints are the same for Enterobacterales but not for P. aeruginosa (susceptible, ≤1 μg/mL) or Acinetobacter spp. (susceptible, ≤1 μg/mL), and the FDA has not published breakpoints for S. maltophilia (18). Cefiderocol breakpoints published by EUCAST are also different from those of the CLSI or FDA. Current EUCAST MIC breakpoints for Enterobacterales and P. aeruginosa are susceptible at ≤2 μg/mL and resistant at >2 μg/mL; non-species-related pharmacokinetic/pharmacodynamic MIC breakpoints for cefiderocol are also susceptible at ≤2 μg/mL and resistant at >2 μg/mL (19). Clearly, nonharmonized breakpoint criteria create perceived differences in susceptibility to cefiderocol, and to other agents, when MICs are interpreted by different MIC breakpoints. Depending upon the interpretive criteria used, isolates of P. aeruginosa with cefiderocol MICs of 2 or 4 μg/mL, for example, may be reported as susceptible, intermediate, or resistant. This is of particular importance for an agent such as cefiderocol because it is intended to be used against Gram-negative pathogens that have elevated MICs for most or all other potential therapeutic agents available.

We conclude that most current (2014 to 2019) clinical isolates of Enterobacterales (99.8%), P. aeruginosa (99.9%), A. baumannii complex (96.0%), and S. maltophilia (98.6%) in North America and Europe are susceptible to cefiderocol by the recently approved CLSI MIC breakpoints (17). Importantly, differences in the annual rates of percent susceptible for cefiderocol from 2014 to 2019 for isolates of Enterobacterales (North America range, 99.6 to 100% susceptible/year; Europe range, 99.3 to 99.9%) and P. aeruginosa (North America range, 99.8 to 100%; Europe range, 99.9 to 100%) were negligible. Annual percent susceptible rates for A. baumannii complex demonstrated sporadic, nondirectional differences (North America range, 97.5 to 100%; Europe range, 90.4 to 97.5%), primarily due to isolates from Russia. Annual percent susceptible rates for S. maltophilia also showed minor, nondirectional fluctuation (North America range, 96.4 to 100%; Europe range, 95.6 to 100%). In vitro susceptibility testing of cefiderocol may be of benefit when cefiderocol is being considered for treatment of patients infected with carbapenem-nonsusceptible, ceftazidime-avibactam-nonsusceptible, or ceftolozane-tazobactam-nonsusceptible isolates of Enterobacterales and P. aeruginosa, carbapenem-nonsusceptible isolates of A. baumannii complex, and MDR isolates of S. maltophilia.

We thank all participating investigators and laboratories who provided isolates for SIDERO-WT studies. SIDERO-WT studies were funded by Shionogi & Co., Ltd., Osaka, Japan, and funding included compensation for manuscript preparation. The sponsor approved the overall study design but the collection and testing of isolates, data analysis, and manuscript preparation were independently performed by IHMA.

J.A.K. is a consultant to IHMA. M.A.H. and D.F.S. are employees of IHMA. M.T. and Y.Y. are employees of Shionogi & Co., Ltd., Osaka, Japan, and R.E. is an employee of Shionogi, Inc., Florham Park, NJ, USA. The IHMA authors do not have personal financial interests in Shionogi & Co., Ltd., or Shionogi, Inc.