Treatment of carbapenem-resistant Pseudomonas aeruginosa infections: a case for cefiderocol

ABSTRACT

Introduction

Carbapenem-resistant (CR) Pseudomonas aeruginosa infections constitute a serious clinical threat globally. Patients are often critically ill and/or immunocompromised. Antibiotic options are limited and are currently centered on beta-lactam–beta-lactamase inhibitor (BL–BLI) combinations and the siderophore cephalosporin cefiderocol.

Areas covered

This article reviews the mechanisms of P. aeruginosa resistance and their potential impact on the activity of current treatment options, along with evidence for the clinical efficacy of BL–BLI combinations in P. aeruginosa infections, some of which specifically target infections due to CR organisms. The preclinical and clinical evidence supporting cefiderocol as a treatment option for P. aeruginosa involving infections is also reviewed.

Expert opinion

Cefiderocol is active against most known P. aeruginosa mechanisms mediating carbapenem resistance. It is stable against different serine- and metallo-beta-lactamases, and, due to its iron channel-dependent uptake mechanism, is not impacted by porin channel loss. Furthermore, the periplasmic level of cefiderocol is not affected by upregulated efflux pumps. The potential for on-treatment resistance development currently appears to be low, although more clinical data are required. Information from surveillance programs, real-world compassionate use, and clinical studies demonstrate that cefiderocol is an important treatment option for CR P. aeruginosa infections.

KEYWORDS:

• Beta-lactam–beta-lactamase inhibitor
• carbapenem resistance
• cefiderocol
• cephalosporin
• CR Pseudomonas aeruginosa
• siderophores
• susceptibility

1. Introduction

Antibiotic-resistant Pseudomonas aeruginosa presents a challenging clinical scenario. P. aeruginosa is able to readily develop resistance to a number of commonly used antibiotics [1,2], including first-line antipseudomonal agents [1,3,4]. As such, P. aeruginosa is one of the most commonly isolated carbapenem-resistant (CR) Gram-negative bacteria encountered in the hospital [5]. Isolates that are multidrug-resistant (MDR), extensively drug-resistant (XDR) [1,5–8] and/or with difficult-to-treat resistance (DTR; exhibiting non-susceptibility to piperacillin–tazobactam, ceftazidime, cefepime, aztreonam, meropenem, imipenem-cilastatin, ciprofloxacin, and levofloxacin) [9] have become a fairly common clinical occurrence. The extent of the problem has been recognized by both the US Centers for Disease Control and Prevention, which considers MDR/CR P. aeruginosa to be a serious threat [10], and the World Health Organization, which listed CR P. aeruginosa among the critical pathogens urgently requiring new antibiotics [11].

The implications of resistance are particularly important given the patient population involved [12,13]. P. aeruginosa infections are frequent in immunocompromised patients, such as those with cancer, and hospitalized patients [10]; therefore, appropriate and timely treatment are essential for avoiding excess morbidity and mortality. Of note, risk factors related to infection severity, especially intensive care unit (ICU) admission, intubation, or invasive devices, are often linked to risk of acquiring MDR P. aeruginosa [2,3,10,14]. Prior carbapenem treatment has also been shown to be a risk factor for the development of CR P. aeruginosa infections [15]. The most frequent infections caused by CR or MDR P. aeruginosa are respiratory tract infections (RTI), nosocomial pneumonia (NP), complicated urinary tract infection (cUTI), and bloodstream infection (BSI) [16–18]. All these infection types caused by CR P. aeruginosa are associated with an approximate 10–35% all-cause mortality rates [16,18,19] as well as with high rates of treatment failure and recurrence [17,18,20].

Despite recent approvals of new antibiotics, there remain relatively few active treatment options for CR P. aeruginosa infections (Figure 1). Choices center on the beta-lactam–beta-lactamase inhibitor (BL–BLI) combinations, such as ceftolozane-tazobactam (C/T), ceftazidime-avibactam (CZA), and imipenem-relebactam (I/R), and the siderophore cephalosporin, cefiderocol. Meropenem–vaborbactam (MVB) may also have a somewhat limited role, with evidence for activity against some strains of MDR/CR P. aeruginosa [21–24]; however, MVB, unlike the other new approved antibiotics, is not included as an option for DTR P. aeruginosa in guidelines from the Infectious Diseases Society of America [9].

Figure 1. Antipseudomonal cephalosporins and their characteristics.

ESBL, extended-spectrum beta-lactamase.

While current data show generally high susceptibility rates of CR P. aeruginosa to newer antibiotics, there is considerable geographical variation (Table 1) [6,7,25–52]. In addition, cross-resistance between C/T and CZA has been noted for carbapenem-non-susceptible strains [28,34,42]. Furthermore, host-dependent risk factors are an important consideration in guiding susceptibility testing, as there have been reports of variable susceptibility rates to C/T and CZA of meropenem-non-susceptible P. aeruginosa isolated from patients with cystic fibrosis (CF) [53].

Table 1. Summary of in vitro susceptibility of ceftazidime–avibactam, ceftolozane–tazobactam, cefiderocol, imipenem–relebactam, and meropenem–vaborbactam against carbapenem-resistant Pseudomonas aeruginosa in different regions’ surveillance programs

	Ceftazidime–avibactam	Ceftolozane–tazobactam	Cefiderocol	Imipenem–relebactam	Meropenem–vaborbactam
North America	57.9% [25]^a	77.7% [36]^b	100% [41]^c 99.8% [42]^a	63.2% [25]^a	NA
USA	74.3% [6]^a 77.8–84.5% [26]^a 89.4% [27]^c 88.5% [28]^a 79.6% [29]^g	71.7% [6]^a 81.2–82.9% [26]^a 80.5% [7]^d 92.4% [27]^c 89.0% [28]^a 88.2% [29]^g 87.8% [37]^a	100% [42]^d 100% [27]^c 98.7% [29]^g	85% [49]^e 82.2% [29]^g	30.9% [29]^g,i
Canada	74.1% [30]^a	92.6% [30]^a	100% [30]^a	NA	NA
Latin America	61.9% [31]^a	66.1% [38]^a 35.9% [36]^b	NA	59.6% [50]^e	NA
Europe	9.1% [32]^f	36.3% [36]^b 9.1% [32]^f Western Europe: 75.2% [39]^a Eastern Europe: 59.2% [39]^a	100% [44]^d 90.9% [32]^f 100% [45]^g 100% [46]^g 96% [47]^g	74.4% [51]^e	NA
Asia	83.6% [33]^a 41% [34]^c 77.3% [35]^g,h	48.3% [36]^b 58.8–69.4% [40]^a 37.9% [34]^c 94.7% [35]^g,h	100% [48]^c	61.2% [52]^e	NA

^aMeropenem non-susceptible; ^bbeta-lactam non-susceptible; ^ccarbapenem non-susceptible; ^dcarbapenem resistant; imipenem non-susceptible; ^fcarbapenemase-producing isolates; ^gmeropenem resistant; ^himipenem resistant; ⁱonly EUCAST breakpoint is available.

EUCAST, European Committee on Antimicrobial Susceptibility Testing; NA, not available.

2. Mechanisms of resistance in carbapenem-resistant P. aeruginosa with impact on susceptibility to beta-lactam–beta-lactamase inhibitor combinations and cefiderocol

Both intrinsic and acquired (through horizontal gene transfer or mutations) antibiotic resistance mechanisms (Table 2) described in P. aeruginosa present a challenge for the selection of antibiotic treatment [3,54–70]. Restriction of the outer membrane permeability is a key factor in intrinsic resistance [54,71]. Constitutive efflux pumps (e.g. MexAB-OprM) contribute to resistance to meropenem and several other classes of antibiotics, including fluoroquinolones and aminoglycosides [54]. As described below, mutation-associated loss or decreased functionality of the key outer membrane porin channel protein OprD [72,73] and overproduction of AmpC, a chromosomal cephalosporinase (or Pseudomonas-derived cephalosporinase [PDC] in P. aeruginosa) [73], can work in concert with the efflux pumps to cause further antibiotic resistance.

Table 2. Mechanisms of resistance contributing to reduced activity or resistance to beta-lactam–beta-lactamase inhibitors and cefiderocol in carbapenem-resistant Pseudomonas aeruginosa.

Mechanism	Ceftazidime-avibactam	Ceftolozane-tazobactam	Cefiderocol	Imipenem-relebactam	Meropenem-vaborbactam
Efflux pump extrusion	Affected [54–56]	Not affected [56]	Not affected [62]	Not affected [55,67]	Affected [57]
Cell entry via OprD porin channel, mutations will decrease expression	Not affected alone [56]	Not affected alone [56]	Not affected [62]	Affected (imipenem) [68] Not affected (I/R, low MIC) [67]	Affected (meropenem, vaborbactam) [57,69,70]
Class C – Constitutive basal AmpC/PDC-mediated hydrolysis	Not affected [54]	Not affected [54]	Not affected [63]	Not affected [67]	Not affected [3]
Class C – De-repressed AmpC/PDC-mediated hydrolysis	Stable [54] Affected (with other mechanism [porin]) [57]	Stable [54] Affected (with other mechanism [porin, efflux]) [54,57]	Stable and not affected [63]	Stable [67]	Possibly stable, no data on Pseudomonas
Class C – Increased AmpC/PDC-mediated hydrolysis via structural change	Affected [54]	Affected [54]	May be affected [64]	Variable [55]	Affected [3]
Class A – ESBLs (SHV, TEM, GES, PER, VEB) mediated hydrolysis	Stable (except PER, GES, VEB) [58,59]	Stable (except PER, GES, VEB) [58,59]	Stable (and modestly affected by PER, VEB) [65]	Stable or modestly affected [59]	Stable [57]
Class A – Carbapenemases (KPC, GES) mediated hydrolysis	Stable [58,59]	Hydrolyzed [58,59]	Stable and not affected [62,66]	Stable or affected [59]	Stable [57]
Class B – MBLs (VIM, IMP, NDM) mediated hydrolysis	Hydrolyzed [59]	Hydrolyzed [59]	Stable and not affected (but higher MICs for NDM isolates) [62,65]	Hydrolyzed [55,59]	Hydrolyzed [57]
Class D – OXAs (chromosomal OXA-50) mediated hydrolysis	Hydrolyzed [60]	Hydrolyzed [60]	No data on Pseudomonas	Hydrolyzed [59]	Hydrolyzed [57]
Class D – OXAs (plasmid OXA-2, OXA-10) mediated hydrolysis	Stable or hydrolyzed [57,61]	Stable or hydrolyzed [57,61]	No data on Pseudomonas	Stable [61,67]	Hydrolyzed [57]
Cross-resistance	Yes, with C/T	Yes, with CZA	May increase MIC but remains susceptible	Yes, with C/T, CZA	Yes, with C/T, CZA

Definitions: Stable: not hydrolyzed; Hydrolyzed: the enzyme has a hydrolytic activity against one of the components of the antibiotic combination or cefiderocol, and activity is reduced; Not affected: any mutation, up- or downregulation has no impact directly on the activity of the antibiotic; Affected: a mutation, up- or downregulation will have a direct impact on the activity of the antibiotic; Variable: a mutation, up- or downregulation may have a direct impact on the activity of the antibiotic.

AmpC, chromosomal cephalosporinase; C/T, ceftolozane–tazobactam; CZA, ceftazidime–avibactam; ESBL, extended-spectrum beta-lactamase; GES, Guiana extended-spectrum beta-lactamase; IMP, imipenemase metallo-beta-lactamase; I/R, imipenem-relebactam; KPC, Klebsiella pneumoniae carbapenemase; MBL, metallo-beta-lactamase; MIC, minimum inhibitory concentration; NDM, New Delhi metallo-beta-lactamase; Opr, outer membrane porin channel; OXA, oxacillinase; PDC, Pseudomonas-derived cephalosporinase; PER, Pseudomonas extended resistance; SHV, sulfhydryl variant of TEM extended-spectrum beta-lactamase; TEM, Temoneira extended-spectrum beta-lactamase; VEB, Vietnamese extended-spectrum beta-lactamase; VIM, Verona integron-encoded metallo-beta-lactamase.

OprD provides entry to P. aeruginosa isolates for several antibiotic classes, including imipenem and meropenem [54,71] but not ertapenem. While oprD mutation in the absence of other resistance mechanisms does not affect the activity of C/T, CZA, and I/R, loss of permeability leads to reduced susceptibility for MVB in CR isolates, although data are scarce [57,70] (Table 2). Porin mutation also restricts imipenem entry into P. aeruginosa, leading to resistance. However, the combination of imipenem with relebactam has been shown to reduce imipenem minimum inhibitory concentrations (MICs) to ≤1 μg/mL in isolates with AmpC hyperproduction, OprD inactivation, and/or efflux pump overexpression [55,67]. Further studies also have shown that relebactam reduced imipenem MICs in isolates producing extended-spectrum beta-lactamases (ESBLs), including Vietnamese extended-spectrum beta-lactamase (VEB), Pseudomonas extended resistance (PER) ESBL, and Guiana extended-spectrum beta-lactamase (GES), but only to levels that remained largely above the susceptibility breakpoint [59]. Due to the distinctive mechanism used by cefiderocol to gain entry into bacterial cells, an action mediated via iron-transport channels, neither porin loss nor upregulated efflux pump have been found to impact on P. aeruginosa susceptibility to this antibiotic [74,75].

Chromosomal constitutive (low expression) AmpC can be induced by a number of antibiotics, including aminopenicillins, some cephalosporins, and carbapenems. In the presence of other resistance mechanisms, AmpC (or PDC) induction and de-repression may also result in resistance to beta-lactam antibiotics other than C/T, CZA, and carbapenems. C/T and CZA are stable against induced or de-repressed AmpC by virtue of ceftolozane and avibactam, respectively (Table 2). During treatment, however, both C/T and CZA can induce resistance through emerging mutations in AmpC/PDC, with the result that CR P. aeruginosa may become cross-resistant to other BL–BLI combinations [64,76,77]. A recent study demonstrated that specific mutations in the AmpC enzyme omega loop can lead to resistance to C/T and CZA, and slightly reduced susceptibility to cefiderocol in the absence of cefiderocol exposure, and increased susceptibility to I/R, a phenomenon known as ‘collateral sensitivity’ [64]. In the study, in contrast to what was observed with C/T and CZA, cefiderocol MIC values rarely changed to non-susceptibility (i.e. >4 µg/mL) [64].

Another form of acquired carbapenem resistance in P. aeruginosa involves the horizontal transfer of genes encoding beta-lactamases, including carbapenemases [14,54]. The relative expression by CR P. aeruginosa of the horizontally acquired ESBLs and carbapenemases shows regional variation [1]. For example, in Colombia, Klebsiella pneumoniae carbapenemase (KPC)-producing isolates occur as frequently as Verona integron-encoded metallo-beta-lactamase (VIM) producers [78]. In Asia, imipenemase-1 carbapenemase (IMP-1) was the most common metallo-beta-lactamase (MBL) in Japan [79], while in India, both VIM and New Delhi MBL (NDM) are frequent carbapenemases in CR P. aeruginosa [80]. In addition, constitutive expression of horizontally transferred genes for non-inducible AmpC enzymes (e.g. FOX [cefoxitin-hydrolyzing class C beta-lactamase] or CMY [cephamycinase]) increases the baseline rate of AmpC-mediated hydrolysis [55].

The use of beta-lactamase inhibitors to restore beta-lactam susceptibility can be impeded by additional P. aeruginosa resistance mechanisms. Relebactam, an irreversible inhibitor of AmpC [55], is not removed by upregulated efflux pumps and so can restore susceptibility to imipenem in AmpC hyperproducers and in those with certain class A enzymes (e.g. KPC). However, susceptibility rates are variable and depend on the expression and/or type of class A enzyme [55,59]. For example, expression by CR P. aeruginosa of either an ESBL or a carbapenemase variant of the class A GES enzyme can reduce susceptibility to I/R [55]. In addition, in vivo evolution of GES beta-lactamases can be driven by CZA treatment and may lead to a switch in the phenotype from carbapenem resistant to susceptible [81]. Vaborbactam is an irreversible, non-beta-lactam, cyclic boronic acid-based inhibitor of serine-beta-lactamases, in particular class A and class C enzymes, mainly in Enterobacterales [82]. Preclinical experiments and surveillance studies with MVB in P. aeruginosa yielded apparently conflicting results. While addition of vaborbactam led to no increase in the in vitro activity, the in vivo bactericidal effects of MVB against certain strains of P. aeruginosa were greater than those of meropenem alone [83]. Based on evidence from Enterobacterales, the failure of vaborbactam to reduce meropenem MICs in P. aeruginosa may be the result of concomitant porin channel mutations, efflux pump upregulation, and/or production of class B carbapenemases (i.e. MBLs) [57,82,84].

Carbapenem resistance in P. aeruginosa is commonly due to the production of MBLs, the prevalence of which varies by hospital and by region. Rates are concerningly high in Singapore [34], India [80,85], Japan [86], and Europe [87,88] but appear to be lower in the USA [89]. MBL-mediated CR P. aeruginosa isolates are generally resistant to treatment with C/T, CZA, I/R, and MVB [34,82,90,91]. Cefiderocol is mostly stable against hydrolysis by MBLs [62,92].

The production of carbapenemases by CR P. aeruginosa, and the type of carbapenemases produced, have considerable therapeutic implications. Rapid diagnostic tests could help to identify patients harboring carbapenemase-producing isolates [93] and so guide appropriate treatment. The need for testing should be guided by local epidemiology [94]. An algorithm to predict P. aeruginosa-associated carbapenemase production has been developed, which incorporates non-susceptibility to carbapenems and cefepime, ceftazidime, and piperacillin-tazobactam (with or without non-susceptibility to C/T or CZA) [95] and also phenotype testing with modified carbapenem inactivation method to confirm the presence of an active carbapenemase [95]. Accurate detection of various enzymes in carbapenemase-positive CR P. aeruginosa has been successful by subsequent rapid diagnostic testing; however, an expanded spectrum of detectable enzymes was needed to uncover the true prevalence of carbapenemases in isolates collected from different regions [94].

In addition, the development of biofilms by P. aeruginosa is an adaptive response mechanism that impedes antibiotic penetration into bacterial cells [2,54,71,96]. Through biofilm formation, P. aeruginosa undergoes phenotypic changes and may become less susceptible or resistant to antibiotics in patients with cystic fibrosis or other chronic lung infections, who therefore often require prolonged and repeated antibiotic treatment [71]. In biofilms, the so-called rugose small colony variants of P. aeruginosa cells are responsible for the excessive production of extracellular matrix elements, which, by creating a thick layer, very efficiently protect these cells from antibiotics allowing them to adapt to the environment and to persist for several weeks [97,98].

3. Treatment of carbapenem-resistant P. aeruginosa infections

The presence of documented clinical risk factors for serious CR/MDR P. aeruginosa infections in hospitalized patients and a knowledge of local susceptibility data corresponding to mechanisms mediating carbapenem resistance and/or history of travel to endemic regions are critical for guiding selection of the most appropriate antibiotic therapy [14,80,94]. Prior hospitalization, ICU stay, tracheal intubation, Acute Physiology and Chronic Health Evaluation II score, diabetes mellitus, and urinary catheter are all associated with MDR P. aeruginosa and/or an increased risk of mortality in MDR P. aeruginosa infections [2,14]. Infection- and treatment-associated risk factors for MDR P. aeruginosa include colonization, prior broad-spectrum antibiotic use, and prior carbapenem use [2,14]. Management decisions should also include consideration of a patient’s travel history, given that there is geographical diversity in the extent and type of carbapenemase production by P. aeruginosa [94].

P. aeruginosa is also a primary causative pathogen of infections in patients with CF and leads to considerable morbidity and mortality [71]. The duration of antibiotic treatment is based on the type of infection and activity of the initial empiric treatment [9,99].

Resistance to aminoglycosides and fluoroquinolones, once the mainstay of treatment and sometimes still active against CR P. aeruginosa, has led to the search for new approaches for the front-line setting. Antibiotic combinations may provide a solution, based on findings of synergy between some antibiotics in in vitro studies [54]. However, this concept is not well established [100–102], and clinical evidence supporting combination use is still scarce despite being widely used in real life [101,103]. It is reasonable to consider empiric antibiotic therapy, based on local susceptibility pattern, to increase the likelihood that at least one agent is active. The approved BL–BLI combinations, CZA, C/T, and I/R, have demonstrated effectiveness in P. aeruginosa infections, including CR forms, in the clinical setting (Table 3) [69,70,104–124]. These agents are recommended as monotherapy for DTR P. aeruginosa once their in vitro activity is confirmed [9]. Because dosing recommendations for these antibiotics are available primarily for adult patients, including elderly patients, management of infections due to CR P. aeruginosa is still highly challenging for certain patient populations such as pediatric patients or during pregnancy.

Table 3. Comparison of recently approved antibiotics with in vitro activity against carbapenem–resistant Pseudomonas aeruginosa.

	Ceftazidime-avibactam	Ceftolozane–tazobactam	Cefiderocol	Imipenem–cilastatin–relebactam	Meropenem–vaborbactam
First approval year	2015 [104,105]	2014 [110,111]	2019 [147,148]	2019 [117,118]	2017 [69,70]
Indications	USA: cUTI, cIAI, HABP, VABP in adult patients and cUTI, cIAI in pediatric patients [104] Europe: cUTI, cIAI, HAP, VAP in adult and pediatric patients [105] Bacteremia associated with cUTI, cIAI, HAP, VAP [105] Infections due to aerobic Gram-negative with limited treatment options [105]	USA: cUTI, including pyelonephritis, cIAI, HABP, VABP in adults [110] Europe: cUTI, acute pyelonephritis, cIAI, HAP, VAP in adults [111]	USA: cUTI, HABP, VABP in adults [147] Europe: Infections due to aerobic Gram-negative bacteria with limited treatment options [148]	USA: HABP, VABP in adult patients, and cUTI and cIAI in adult patients with limited or no alternative treatment options [117] Europe: HAP, VAP [118] Bacteremia associated with HAP, VAP [118] Infections due to aerobic Gram-negative bacteria with limited treatment options [118]	USA: cUTI in adult patients [69] Europe: HAP, VAP, cUTI, cIAI [70] Bacteremia associated with HAP, VAP, cUTI, cIAI [70] Infections due to aerobic Gram-negative bacteria with limited treatment options [70]
Dosing	2 g/0.5 g, q8h, 2-h infusion in adult patients 50 mg/kg/12.5 mg/kg to a maximum of 2 g/0.5 g, q8h, 2-h infusion in pediatric patients, or 40 mg/kg/10 mg/kg in pediatric patients aged 3–<6 months	1 g/0.5 g, q8h, 1-h infusion (for cUTI, cIAI) 2 g/1 g, q8h, 1-h infusion (for pneumonia) Pediatric dosing is not yet available	2 g, q8h, 3-h infusion Pediatric dosing is not yet available	500 mg/500 mg/250 mg, q6h, 30-min infusion Pediatric dosing is not yet available	2 g/2 g, q8h, 3-h infusion Pediatric dosing is not yet available
Activity against certain CR mechanisms	AmpC, Class A, Class D	AmpC	AmpC, Class A, Class B, Class C, Class D, Porin, Efflux	AmpC, Class A, Efflux	AmpC, Class A
Clinical efficacy in clinical studies	RECAPTURE (Phase III): cUTI/pyelonephritis [106] Among 810 patients, 38 had P. aeruginosa as a baseline pathogen 1° endpoints: patient-assessed symptomatic resolution at Day 5: 70.2% CZA vs 66.2% doripenem; patient-assessed symptomatic resolution plus favorable per patient microbiological response: 71.2% CZA vs 64.5% doripenem REPRISE (Phase III): cUTI or cIAI involving CAZ-resistant Enterobacterales or P. aeruginosa [107] Of 302 evaluable patients, 21 had CAZ-resistant P. aeruginosa isolate 1° endpoint: clinical cure at TOC: 91% CZA vs 91% BAT REPROVE (Phase III): Nosocomial pneumonia [108] Among 355 patients, 105 had P. aeruginosa infection, of which 25 were CAZ non-susceptible 1° endpoint: clinical cure at TOC: 68.8% CZA vs 73% meropenem Phase III: cIAI [109] Among 823 patients, 6 had CAZ-resistant P. aeruginosa 1° endpoint: clinical cure at TOC: 81.6% CZA + metronidazole vs 85.1% meropenem	Phase II: cIAI [112] Among 121 patients in the modified ITT population, 7 patients had P. aeruginosa infection eligible for the microbiological outcome evaluation 1° endpoints: clinical response at TOC: 83.6% TOL-TAZ + metronidazole vs 96.0% meropenem Microbiological success in infections caused by P. aeruginosa: 100% TOL-TAZ + metronidazole vs 100% meropenem ASPECT-cIAI (Phase III) cIAI [113] Among 993 randomized patients, 53 patients had P. aeruginosa infection in the microbiologically evaluable population 1° endpoint: clinical cure at TOC: 83.0% TOL-TAZ + metronidazole vs 87.3% meropenem Clinical cure in infections caused by P. aeruginosa: 100% TOL-TAZ + metronidazole vs 93.1% meropenem ASPECT-cUTI (Phase III): cUTI/pyelonephritis [114] Among 800 patients, 23 had P. aeruginosa infection 1° endpoint: composite cure: 76.9% C/T vs 68.4% levofloxacin ASPECT-NP (Phase III): nosocomial pneumonia [115] Among 511 patients in the microbiological ITT population, 4/127 with P. aeruginosa as a baseline pathogen had an isolate resistant to C/T 1° endpoint (ITT population [n = 726]): 28-day ACM: 24% C/T vs 25.3% meropenem Retrospective: 21 patients with MDR P. aeruginosa infections [116] C/T success rate: 71% C/T resistance emerged in 14%	APEKS-cUTI (Phase II): cUTI/pyelonephritis [149] Among 371 patients, 23 had P. aeruginosa as a baseline pathogen 1° endpoint: composite clinical and microbiological outcomes at TOC: 73% CFDC vs 55% I/C APEKS-NP (Phase III): HABP, HCABP, VABP [150] Among 292 patients, 48 had P. aeruginosa as a baseline pathogen 1° endpoint: Day-14 ACM: 12.4% CFDC vs 11.6% meropenem CREDIBLE-CR (Phase III): CR infections [151] Among 118 patients, 22 had CR P. aeruginosa as a baseline pathogen 1° endpoint: clinical cure at TOC for nosocomial pneumonia (50% CFDC vs 53% BAT) and BSI/sepsis (43% CFDC vs 43% BAT); microbiological eradication for cUTI (53% CFDC vs 20% BAT)	Phase II: cUTI [119] Among 230 patients, 50% of 16 P. aeruginosa isolates at baseline were imipenem resistant 1° endpoint: microbiological response at discontinuation of IV therapy: 95.5–98.6% I/C + relebactam vs 98.7% I/C + placebo RESTORE-IMI 1 (Phase III): Imipenem-non-susceptible infections in hospitalized patients [120] P. aeruginosa was the most frequent qualifying pathogen (77%) 1° endpoint: favorable overall response: 71% I/R vs 70% colistin+imipenem Favorable overall response against P. aeruginosa was observed in 13/16 (81%) of I/R and 5/8 (63%) of colistin+imipenem patients RESTORE-IMI 2 (Phase III): HAP/VAP [121] Among 531 patients in the mITT population, 18.9% had P. aeruginosa as a baseline pathogen 1° endpoint: 28-day ACM: 15.9% I/R vs 21/3% PIP/TAZ	TANGO-I (Phase III): cUTI/pyelonephritis [122] Among 550 patients, there were none with P. aeruginosa at baseline 1° FDA endpoint: composite of clinical and microbiological outcomes at EOT: 98.4% MVB vs 94.0% PIP/TAZ 1° EMA endpoint: microbiological eradication at TOC: 66.7% MVB vs 57.7% PIP/TAZ TANGO-II (Phase III): cUTI/pyelonephritis, cIAI, HAP/VAP, bacteremia [123] Among 77 patients, there were none with P. aeruginosa at baseline Overall clinical cure at TOC: 59.4% MVB vs 26.7% PIP/TAZ Day 28 ACM for nosocomial pneumonia, bacteremia: 15.6% MVB vs 33.3% PIP/TAZ Retrospective, observational, cohort MDR Gram-negative infections [124] Among 126 patients, 11 had P. aeruginosa at baseline 1° endpoint: Day-30 ACM was 18.3% overall; 0% in P. aeruginosa infections and 19.5% in non-P. aeruginosa infections 11.9% had recurrence within 30 days
Safety/AEs in clinical studies	4 patients had an SAE possibly related to CZA (diarrhea, ACS, subacute hepatic failure, liver function test abnormal) [108] Clostridioides difficile colitis (3 patients with CZA vs 0 with doripenem) [106]	Incidence of SAEs were higher in TOL-TAZ arm and were related to treatment failure, the underlying IAI and surgical procedure; none of the SAEs were related to study drug [112] Two drug-related SAEs of Clostridioides difficile infection related to C/T [113] Two SAEs of Clostridioides difficile infection related to C/T [114]	One CFDC-related case of Clostridioides difficile colitis [149] Liver-related TEAEs, usually in patients with confounding factors; none were related to treatment [151] Adverse events leading to discontinuation included diarrhea, drug hypersensitivity, increased hepatic enzymes in patients with cUTI [147], and elevated liver tests in patients with pneumonia [147].	Two cases of increased liver enzymes (one drug related, one non-drug related) [119]	Lower rate of renal-related TEAEs or acute renal failure with MVB [123] 1 case of CDAD in TANGO-2 [82,123] 2 cases of nephrotoxicity, 1 case of rash, 1 case of hepatotoxicity [124]

ACM, all-cause mortality; ACS, acute coronary syndrome; BAT, best available therapy; BSI, bloodstream infection; CAZ, ceftazidime; CDAD, Clostridium difficile-associated diarrhea; CFDC, cefiderocol; cIAI, complicated intra-abdominal infection; CR, carbapenem resistant; C/T, ceftolozane–tazobactam; cUTI, complicated urinary tract infection; CZA, ceftazidime–avibactam; EMA, European Medicines Agency; EOT, end of treatment; FDA, US Food and Drug Administration; HA(B)P, hospital-acquired (bacterial) pneumonia; HCABP, healthcare-associated bacterial pneumonia; I/C, imipenem–cilastatin; I/R, imipenem–relebactam; ITT, intention-to-treat; IV, intravenous; MDR, multidrug resistant; mITT, modified intention-to-treat; MVB, meropenem–vaborbactam; PIP/TAZ, piperacillin–tazobactam; q6h, every 6 h; q8h, every 8 h; SAE, serious adverse event; TEAE, treatment-emergent adverse event; TOC, test of cure; TOL-TAZ, ceftolozane–tazobactam; VA(B)P, ventilator-associated (bacterial) pneumonia.

CZA was active against 72% of CR P. aeruginosa isolates from the ERACE-PA Global Surveillance Program (2019–2021), and against 91% of carbapenemase-negative isolates [94]. In the INFORM Global Surveillance Program (2012–2015) involving isolates from Asia-Pacific countries, susceptibility of MBL-negative meropenem-non-susceptible and MDR P. aeruginosa isolates was 83.6% and 68.2%, respectively [33]. The clinical efficacy of CZA has been demonstrated in cUTI, complicated intra-abdominal infection (cIAI), and NP in the multicenter, Phase III RECAPTURE [106], REPRISE [107], REPROVE [108], and RECLAIM [109] trials, including patients with P. aeruginosa infections, some with CR forms (Table 3). In a pooled analysis of the Phase III trial program, among 95 patients with ≥1 MDR P. aeruginosa isolate, the favorable microbiological response rate with CZA was 57.1% [125].

For C/T, susceptibility rates in CR P. aeruginosa overall and carbapenemase-negative CR P. aeruginosa isolates were 63% and 88%, respectively, in the ERACE-PA Global Surveillance Program (2019–2021) [94]. The multinational SIDERO surveillance program recently reported that the average susceptibility rate for C/T in P. aeruginosa isolates was 76%, reaching 90% among US isolates, thereby indicating a higher proportion of resistance in Europe [42]. The Phase III ASPECT trials investigating C/T in cIAI, cUTI and NP included 5.3%, 3%, and 25% of patients, respectively, with P. aeruginosa infections [113–115]; efficacy of C/T was similar to that of comparator agents. In a retrospective study of 21 patients with MDR P. aeruginosa infections (86% RTIs), the success rate of C/T was 71%, but resistance emerged in 3 patients (14%) [116].

For I/R, data from the SMART Surveillance Europe database (2015–2017) reported that among P. aeruginosa isolates from IAI and UTI sources, 94.4% and 93.0%, respectively, were susceptible to I/R; the rate was 79.8% for MDR isolates from IAI and UTI combined [51]. In the RESTORE-IMI 1 trial, 77% of imipenem-non-susceptible infections were due to P. aeruginosa [120]; a favorable response against P. aeruginosa was observed in 13/16 patients (81%) receiving I/R. In the RESTORE-IMI 2 trial in hospital-acquired pneumonia/ventilator-associated pneumonia, 18.9% of 531 patients had P. aeruginosa as a baseline pathogen; 28-day all-cause mortality was 33.3% among a small number of patients with P. aeruginosa treated with I/R versus 12.0% among those who received piperacillin/tazobactam [121].

For MVB, limited clinical data are available; none of the patients involved in the Phase III TANGO-I and TANGO-II trials in cUTI/pyelonephritis had P. aeruginosa infections at baseline [122,123]. In a real-world cohort of patients receiving MVB for a variety of Gram-negative infections, none of the eight patients in whom MVB was used to primarily target P. aeruginosa died by Day 30 [124].

4. Cefiderocol for the treatment of carbapenem-resistant P. aeruginosa infections

4.1. In vitro activity

Cefiderocol is a siderophore cephalosporin (Figure 1), a beta-lactam conjugated with a catechol group, and therefore acts as an iron chelator [74]. It binds penicillin-binding proteins and inhibits cell wall synthesis in Gram-negative bacteria [74]. Cefiderocol demonstrates bactericidal activity against Gram-negative bacteria but lacks activity against anaerobic bacteria and Gram-positive bacteria [62,74,75].

Cefiderocol enters bacterial cells through iron-transport channels via an active uptake mechanism [74,75]. Bacterial iron uptake is facilitated by specific transport systems including siderophore receptors in the outer membrane, TonB-ExbBD energy transducing proteins, and inner membrane iron transport channels [126]. Two types of natural siderophores, pyoverdines and pyochelin, are produced by P. aeruginosa that facilitate iron uptake through iron transporters [127]. Cefiderocol exploits two different routes: PiuA and PiuD are utilized in the absence of pyochelin, and FptA in the presence of pyochelin [128]. A deficiency of PiuA or overproduction of pyoverdine due to pvdS mutation has been shown to lead to an 8- to 16-fold increase in the cefiderocol MIC [128,129]. However, resistance has not been observed to emerge under human pharmacokinetic exposure in in vitro chemostat models and in vivo murine thigh infection models [130].

Cefiderocol is able to counter all mechanisms of carbapenem resistance. Cefiderocol is a single molecule and not a BL–BLI fixed combination antibiotic [74]. Its stability against hydrolysis by beta-lactamases, including carbapenemases, is intrinsic to the molecular structure and is linked to the various side chains, including the catechol moiety responsible for iron binding [74,131]. It is mostly stable against serine-beta-lactamases (classes A, C, and D) and against MBLs [62,63,92]. An exception to this is class A PER or VEB ESBL, which can be expressed by P. aeruginosa [65]. PER-producing P. aeruginosa has a higher cefiderocol MIC than P. aeruginosa strains expressing other types of beta-lactamases (e.g. MBL, AmpC) [62,65]. According to a UK study, at ≤4 μg/mL, 73.3% of 15 PER-producing isolates and 90% of 10 VEB-producing isolates were susceptible to cefiderocol [65]. Cefiderocol also maintains activity against P. aeruginosa strains that overproduce AmpC [63]. In addition, because of the iron-transport channel-related mode of bacterial cell entry used by cefiderocol, porin channel loss and/or efflux pump upregulation have minimal impact on its antibacterial activity [62].

Cefiderocol has also been reported to reduce P. aeruginosa-associated biofilm formation and its mass, an effect that tends to show concentration and time dependency [132]. This effect may be partially explained by increased synthesis of natural siderophores and possible upregulation of siderophore transporters by P. aeruginosa during biofilm formation in order to meet increased demand for iron [132,133]. However, further investigation is needed to confirm the activity of cefiderocol in suppressing the growth of small colony variant forms of P. aeruginosa.

Cefiderocol has demonstrated excellent in vitro activity against clinical isolates of P. aeruginosa in the SIDERO-WT and SIDERO-CR surveillance studies, including against DTR strains and strains resistant/non-susceptible to C/T, CZA, and I/R [42,134–138]. Activity has been shown against respiratory, bloodstream, and urinary tract infection isolates [27,139]. Cefiderocol has also demonstrated in vitro activity against MDR P. aeruginosa strains isolated from cancer patients (i.e. 97.5% were inhibited at ≤4 μg/mL) [140] and from ICU patients (i.e. 100% were inhibited at ≤1 μg/mL) [30], and against a pandrug-resistant P. aeruginosa clinical isolate producing MBLs [141].

Susceptibility testing is conducted using either broth microdilution with iron-depleted cation-adjusted Mueller–Hinton broth or using 30 µg disks on standard Mueller–Hinton agar plates without the need to adjust iron concentrations in the agar; these methods are endorsed by both Clinical and Laboratory Standards Institute (CLSI) and European Committee on Antimicrobial Susceptibility Testing committees [142]. Gradient diffusion strips, currently only validated for P. aeruginosa, are also commercially available. For disk diffusion and gradient diffusion testing, unlike broth microdilution, iron depletion of media is not required, presumably because iron is bound in the medium, mimicking an iron-depleted status.

4.2. Preclinical models

The in vivo effectiveness and pharmacodynamic target of cefiderocol have been investigated in several animal models of P. aeruginosa infection [143–146]. In neutropenic murine models of P. aeruginosa thigh and lung infection, including CR strains, use of cefiderocol resulted in stasis with ~63% fraction of time over the MIC (fT>MIC) and a 1-log reduction with ~70–72% fT>MIC [144]. The in vivo effectiveness was independent of carbapenemase production [144]. In addition, among 20 P. aeruginosa strains with MICs of ≤4 µg/mL, 85% showed bacterial stasis or ≥1-log growth reduction compared with baseline inoculum in a murine thigh infection model using humanized drug exposure [145]. The in vivo effectiveness was also shown against CR P. aeruginosa using humanized dosing of cefiderocol in a rat model of pneumonia [143]. Notably, unlike the observations with previous investigational siderophore antibiotics (i.e. MB-1, SMC-3176), the in vivo efficacy of cefiderocol has been shown to correlate with its in vitro activity [146].

4.3. Clinical data in carbapenem-resistant Pseudomonas infections

Cefiderocol currently has regulatory approval for the treatment of cUTI and NP in the USA [147], and for infections caused by susceptible pathogens with limited treatment options in Europe [148] (Table 3). Of note, cefiderocol dose adjustment is required for patients with moderate or severe renal impairment; and for patients with augmented renal clearance, the recommendation is to increase the frequency of the infusions (i.e. 2 g, once every 6 h) [147,148]. Approval for cUTI and NP was based on the Phase II APEKS-cUTI study [149] and the Phase III APEKS-NP study, respectively [150]. These two randomized, double-blind, non-inferiority studies enrolled patients with P. aeruginosa infections, which were primarily carbapenem-susceptible isolates, and only a low proportion were CR isolates. However, patients in both studies were at a high risk of being infected with MDR pathogens and dosing of the comparator agent was selected to address MDR Gram-negative pathogens [149,150].

In both studies, comparable clinical and microbiological outcomes were reported between treatment arms for P. aeruginosa infections [149,150]. In patients with P. aeruginosa infections in the APEKS-cUTI study, the rates for the composite of clinical and microbiological outcomes at test of cure (TOC) (which was the primary endpoint in the main population) were 46.7% (7/15) for cefiderocol and 50% (2/4) for imipenem-cilastatin (treatment difference: –3.33), and clinical cure at TOC was around 75% in both arms (cefiderocol: 73.3 [11/15], imipenem-cilastatin: 75.0% [3/4], treatment difference: –1.67) [149]. Cefiderocol MIC₉₀ was 0.25 µg/mL in both treatment arms. In the APEKS-NP study, Day 14 all-cause mortality rates (the primary endpoint in the main population) of 8% (2/24) and 13% (3/23; treatment difference: –4.7; 95% confidence interval [CI]: –22.4; 12.9) were reported with cefiderocol and high-dose, extended-infusion meropenem, respectively, in patients with NP caused by P. aeruginosa [150]. At Day 28, the corresponding all-cause mortality rates were 8% (2/24) and 26% (6/23; treatment difference: –17.8; 95% CI: –38.8; 3.3), respectively. At TOC, the clinical cure rates were 67% (16/24) for cefiderocol and 71% (17/24) for meropenem (treatment difference: –4.2; 95% CI: –30.4; 22.0), and the microbiological eradication rates were 38% (9/24) and 46% (11/24; treatment difference: –8.3; 95% CI: –36.1; 19.5), respectively. Follow-up clinical cure rates were high for both cefiderocol (58%) and meropenem (54%). Cefiderocol MIC₉₀ was 0.5 µg/mL at baseline [150]. The strength of these two clinical studies is that cefiderocol treatment was administered as monotherapy; thus, efficacy of this new agent is not confounded by the activity of another antibiotic.

Probably the most relevant clinical trial data were generated in the CREDIBLE-CR study, which was conducted in patients with serious CR infections and formed the basis of the regulatory approval of cefiderocol for the treatment of patients with infections caused by susceptible pathogens with limited treatment options [151]. The study enrolled patients with a variety of infections, including NP, BSI/sepsis, and cUTI. There was only a very limited number of general exclusion criteria, but diagnosis-specific inclusion and exclusion criteria were also applicable [151]. The primary endpoint was clinical cure rate in patients with NP and BSI/sepsis, and microbiological eradication rate in patients with cUTI. In CR P. aeruginosa infections, the clinical cure rate at TOC was 58% in the cefiderocol arm and 50% in the best-available therapy (BAT) arm. The microbiological eradication rate against CR Pseudomonas was low (1/12 patients), and 7/12 of the isolates had indeterminate responses (mainly due to missing samples) [151]. The cefiderocol MIC₉₀ was 2 µg/mL in the cefiderocol arm, and the highest MIC was 4 µg/mL. A post-hoc analysis of the CREDIBLE-CR study reported that clinical cure at TOC was achieved in all four patients with CR P. aeruginosa isolates producing an MBL enzyme and in seven of 10 patients with OprD deficiency; however, the microbiological findings were confounded by the high indeterminate microbiological response rate (e.g. missing samples, additional antibiotics) [151,152]. All-cause mortality rates at the end of study in the safety population of CREDIBLE-CR were similar between treatment arms in primarily CR P. aeruginosa infections without Acinetobacter spp. co-infection (18% and 18%) and were increased with cefiderocol treatment in polymicrobial infections with Acinetobacter spp. (35% and 17%). Patients who had Acinetobacter spp. at baseline were sicker, older, more frequently in the ICU, and more frequently had prior/ongoing shock, all of which might have contributed to the increased mortality in the cefiderocol arm, including the patients with CR P. aeruginosa co-infections [151].

Clinical evidence is also available for patients who received cefiderocol in a compassionate use setting (Table 4). Patients had a variety of DTR/CR P. aeruginosa infections, including aortic valve endocarditis, osteomyelitis, surgical site infection (wound infection), transplant patient (portal prosthesis), IAI, RTIs, BSI, esophageal-pleural fistula empyema, and meningitis, and had no alternative treatment options. Many of the infections were polymicrobial or patients were colonized with another pathogen (Table 4) [153–167]. Among 29 patients with P. aeruginosa isolates that had cefiderocol MICs up to 4 µg/mL or susceptibility confirmed by disk zone diameter, 24 receiving cefiderocol responded to treatment. Among these 24 responders, 14 patients received cefiderocol in combination therapy and 10 patients received monotherapy. Six patients had isolates with cefiderocol MICs ≥8 µg/mL, some of which belonged to the hypervirulent ST type ST357, and did not respond to treatment [158,161]; cefiderocol was discontinued in one patient due to non-susceptibility [165]. In a series of 46 patients in the compassionate use program, which included a few patients with CF, clinical response to cefiderocol was seen in 67% of patients with cefiderocol MIC values of ≤4 µg/mL (CLSI susceptibility breakpoint) [168]. Response to treatment was observed for >80% of patients with CR P. aeruginosa isolates with cefiderocol MIC >1 µg/mL (i.e. US Food and Drug Administration susceptibility breakpoint). The response rate was similar for the CR P. aeruginosa isolates with cefiderocol MIC 2–4 and ≥8 µg/mL [168]. Among the 95 CR P. aeruginosa isolates, of which 77% were not susceptible to C/T and 69% were not susceptible to CZA, five isolates had cefiderocol MIC >4 µg/mL [168]. These data highlight the cross-resistance between C/T and CZA and maintained cefiderocol susceptibility in a large proportion of CR P. aeruginosa, for which cefiderocol had been an effective choice for treatment [168].

Table 4. Compassionate use of cefiderocol in extensively drug-resistant Pseudomonas aeruginosa infections

Infection site (Co-infecting or colonizing pathogen(s))	Duration of CFDC treatment	CFDC therapy	Adverse events	Disk zone diameter or MIC (μg/mL)/Outcome	Reference
Endocarditis (Rectal colonization with K. pneumoniae and A. baumannii)	4 weeks	Combination with meropenem, colistin	Neutropenia	Disk/Successful eradication	153
Intra-abdominal abscess (VRE and E. coli)	28 days	Monotherapy	None	0.12/Clinical cure	154
Implant-associated osteomyelitis (ESBL K. pneumoniae)	14 weeks	Monotherapy	None	4/Cure, no amputation	155
Osteomyelitis and wound infection (A. baumannii and E. cloacae)	2 weeks	Combination with CZA, colistin	Minor drop in neutrophils	NA/Clinical cure	156
Surgical site infection (No other organism)	14 days (+ fosfomycin)	Combination with fosfomycin	None	0.5/Clinical cure	157
Pancreatic abscess (diabetic foot, BSI, respiratory colonization) (S. aureus, K. pneumoniae)	6 weeks	Combination with metronidazole	Emergence of resistance	S–>R (8 μg/mL)/ Death	158
Empyema (Candida albicans)	3 weeks	Combination with fluconazole	None	0.25/0.5/1^a/Cure + Persistence, relapse	159
Bacteremia/ portal prosthesis (No other organism)	6 weeks	Combination with colistin, and followed by monotherapy	None	Disk/Clinical cure	160
Various (9 patients) (1 case with K. pneumoniae)	NA	Monotherapy or in combination with colistin	Thrombocytopenia (1 patient)	2–4 (4 isolates)/Cure 3 patients 8–>32 (5 isolates)/Failure 5 patients	161
BSI (No other organism)	7 days	Monotherapy	None	‘22 mm’/Cure + discharged	162
Wound infection following orthopedic surgery (lumbar, metalware) (XDR A. baumannii)	32 days	Monotherapy, followed by ciprofloxacin PO	Acute interstitial nephritis Eosinophilia	0.5/Clinical cure	163
Osteomyelitis, infected wound (M. morganii, S. epidermidis)	28 days	Combination with ciprofloxacin	None	S/Clinical cure	164
Sepsis, bacteremia (VRE [E. faecium])	4 days	Combination with metronidazole, polymyxin B	None reported	Discontinued due to non-susceptibility	165
Empyema	4 weeks	Combination with colistin and fosfomycin	None reported	0.5/Clinical cure	166
Various (17 patients) (1 case with PDR A. baumannii, 1 case with S. maltophilia, 1 case with K. pneumoniae)	NA	Monotherapy for 3 patients or combination therapy for 14 patients	One case of encephalopathy	0.12–≤2 (14 isolates)/Cure 11 patients NA (3 isolates)/Cure 1 patient	167

^aPleural fluid isolate (Day 33)/morphological variant 1 (tracheal aspirate; Day 76)/morphological variant 2 (tracheal aspirate; Day 76).

BSI, bloodstream infection; CFDC, cefiderocol; CZA, ceftazidime–avibactam; ESBL, extended-spectrum beta-lactamase; MIC, minimum inhibitory concentration; NA, not available; PDR, pan-drug resistant; PO, oral; R, resistant; S, susceptible; VRE, vancomycin–resistant Enterococcus; XDR, extensively drug resistant.

There are as yet few data detailing development by P. aeruginosa of resistance to cefiderocol. Among 41 P. aeruginosa clinical isolates derived from patients in the APEKS-NP (n = 24) and CREDIBLE-CR (n = 17) studies, three demonstrated a ≥4-fold increase in cefiderocol MIC with treatment [169]. One of these isolates emerged as resistant and two remained susceptible according to CLSI susceptibility breakpoints [169]. In a real-world clinical-use setting, cefiderocol resistance was reported in two patients following cefiderocol treatment [158,159].

5. Conclusions

Among the limited treatment options for CR P. aeruginosa, the most recently approved BL–BLI inhibitor combinations display a high rate of in vitro activity and favorable efficacy and safety profile in clinical studies; however, resistance to all agents has been reported. As a single agent without a beta-lactamase inhibitor, cefiderocol was developed to overcome all mechanisms of carbapenem resistance, including those that have been seen occurring in CR P. aeruginosa in surveillance studies. Thus, cefiderocol, with its potent activity, is a promising treatment choice against CR P. aeruginosa, with proven efficacy in clinical studies and case reports in patients with challenging infections, such as NP or BSI.

6. Expert opinion

Selection of an appropriate antibiotic for the treatment of CR P. aeruginosa from the limited treatment options is a challenging task for the physician, the microbiologist, or the pharmacist given the variety of resistance mechanisms. Risk factors for being infected by P. aeruginosa are well known [14]; however, host factors, prior treatment failure, local antibiogram, travel history, diversity of carbapenem resistance mechanisms, and susceptibility are equally important factors to consider when CR strains are confirmed. Delayed and/or inappropriate antibiotic treatment for septic shock and/or ICU patients with CR infections increases the risk of mortality [11,170,171]; every hour delay in receiving appropriate antibiotic treatment increases the patient’s mortality risk by 7.6% [170]. Outcomes may depend on the infection site, too; UTIs require a shorter treatment period and are associated with a lower mortality risk, whereas BSI and NP carry a greater risk of mortality [99,108,114,115,149–151].

The species and its antibiotic susceptibility (resistance) profile are not known when empiric antibiotic treatment is initiated, although for suspected P. aeruginosa infection, coverage with two antipseudomonal agents is reasonable when resistance risk according to epidemiological situation and patients’ characteristics is high. Many first-line agents can induce resistance even to some of the newer antibiotics, suggesting that careful consideration of available empiric antibiotics is required prior to treatment. Prevention of further resistance development is one of the primary objectives of combination therapy for P. aeruginosa infections [101,102]; however, only a limited number of agents from different classes have proved to act synergistically in vitro, hampering their use in combination. Current treatment trends, at least in Europe, show that polymyxin, either as monotherapy or in combination therapy, is frequently prescribed for CR P. aeruginosa infections [101], despite limited data on clinical evidence supporting combination [101]. Currently, there is no evidence that cefiderocol and other antibiotics have synergistic effects against CR P. aeruginosa because isolates with high cefiderocol MIC values are rare [42,137]. Despite lack of evidence, cefiderocol was administered in combination with other antimicrobials for many patients in the compassionate use program, who received cefiderocol as a last resort treatment because no alternative treatment option was available [168].

Importantly, CR P. aeruginosa has limited treatment options, and cross resistance is already known for the recently approved BL–BLI combinations (e.g. C/T, CZA, I/R, MVB) due to their similar target mechanisms of resistance [29,42,135]. In contrast, cefiderocol has advantages over these antibiotics in its stability against hydrolysis by various beta-lactamases, including many of the ESBLs and metallo-carbapenemases, as well as against resistance due to porin channel loss or upregulated efflux pumps, covering all categories of CR mechanisms in P. aeruginosa. It is less prone to resistance development during treatment, although further relevant clinical data are still required to confirm this observation.

Cefiderocol showed the highest susceptibility rates among the newer treatment options against CR P. aeruginosa in the SIDERO multinational surveillance program, which collected isolates over a 5-year period [24,29,42,137]. In SIDERO, less than 0.1% of P. aeruginosa isolates had cefiderocol MIC >4 µg/mL, and susceptibility rates among P. aeruginosa isolates which were non-susceptible to C/T and CZA were 99.8% and 100%, respectively [42,137]. Among these C/T-non-susceptible and CZA-non-susceptible isolates 85–88% were meropenem non-susceptible, suggesting a very limited number of treatment options [137]. For such cases, cefiderocol could be considered as an active treatment option [9]. In the study by Mushtaq et al. [65], 13.5% of P. aeruginosa isolates demonstrated cefiderocol MIC >4 µg/mL in a special collection of unusual resistant isolates enriched for PER and MBL producers gathered over a 10-year period by the Public Health England Antimicrobial Resistance and HealthCare Associated Infections Reference Unit. To date, all surveillance reports suggest high-level and sustained susceptibility to cefiderocol, strengthening its potential recommendation for routine clinical use against CR P. aeruginosa. Susceptibility testing available for use in clinical microbiology laboratories (recommended) or through reference laboratories could guide recommendations for treatment of individual patients.

One relevant threat among carbapenemase-producing CR P. aeruginosa is the expression of an MBL enzyme, against which there are only two active antibiotics: cefiderocol in monotherapy and aztreonam in combination with a beta-lactamase inhibitor of the diazabicyclooctane group, such as avibactam. MBL-producing CR P. aeruginosa is spreading in some regions [94], and new MBL enzymes may emerge in CR P. aeruginosa. It has been recognized that MBL-producing CR P. aeruginosa infections can have a faster onset of illness and faster progression to death than non-MBL-producing CR P. aeruginosa infections [14,172]. Thus, cefiderocol is a reasonable treatment choice in regions where MBL-producing CR P. aeruginosa is endemic, as none of the other treatment options have coverage against such strains as single agents.

Clinical data with cefiderocol treatment, which was utilized as monotherapy in clinical trials [149–151] and often in combination therapy in the compassionate use program [168] and in case reports (Table 4), have been promising. Cefiderocol was well tolerated in clinical studies. Adverse events reported included events previously associated with the cephalosporin class, such as Clostridioides difficile colitis, diarrhea, rash, or increased liver enzymes (Table 3). Cefiderocol showed similar efficacy to active comparator agents in clinical studies. The APEKS-cUTI and APEKS-NP studies enrolled patients with high risk of being infected by MDR Gram-negative bacteria, including P. aeruginosa. The non-inferiority of cefiderocol to high-dose, extended-infusion meropenem treatment in patients with NP suggests that cefiderocol has sufficient lung penetration to be effective against this infection. All-cause mortality in carbapenem-susceptible P. aeruginosa infections treated with cefiderocol was similar or lower than that of high-dose, extended-infusion meropenem in APEKS-NP, and similar between cefiderocol and best available therapy in CREDIBLE-CR for patients with CR P. aeruginosa without Acinetobacter spp. co-infection. In the subset of patients with Acinetobacter spp. co-infection, an imbalance in the underlying risk factors at randomization was observed, increasing the overall mortality risk. Therefore, more clinical data specific to CR P. aeruginosa infections are warranted to confirm the utility of cefiderocol against this challenging pathogen.

Overall, surveillance programs, real-world compassionate use data, and the CREDIBLE-CR study data demonstrated that cefiderocol is a valuable treatment option, without being restricted by the mechanism of carbapenem resistance, for CR P. aeruginosa infections. These considerations have been reflected in recent clinical practice guidelines or expert recommendations for the treatment of challenging, drug-resistant pathogens, including CR P. aeruginosa [9,99,173].

Article highlights

• Carbapenem-resistant (CR) Pseudomonas aeruginosa continues to be a challenging problem for infectious disease specialists and other physicians due to limited treatment options because of a variety of resistance mechanisms, which often co-exist in this species.

• CR P. aeruginosa has developed multiple mechanisms that confer resistance to different beta-lactam antibiotics, including first-line treatment extended-spectrum, antipseudomonal penicillins, cephalosporins, and carbapenems, and resistance often extends to other classes as well, such as fluoroquinolones, aminoglycosides, and occasionally polymyxins.

• The recent approvals of beta-lactam–beta-lactamase inhibitor (BL–BLI) combinations (e.g. ceftolozane–tazobactam, ceftazidime–avibactam, imipenem–relebactam, and to a lesser extent, meropenem–vaborbactam) offer solutions in the management of CR P. aeruginosa infections with clear evidence of their clinical efficacy; however, emerging resistance and cross-resistance have been reported.

• Cefiderocol, a siderophore cephalosporin, with a distinct mechanism of cell entry via iron transport channels, can bypass many of the known beta-lactam resistance mechanisms. Moreover, among isolates resistant to the abovementioned BL–BLI combinations, cefiderocol has displayed a high level of in vitro activity.

• Cefiderocol is stable against nearly all beta-lactamases, including class A Klebsiella pneumoniae carbapenemase, class B metallo-beta-lactamases, class C AmpC, and class D oxacillinase enzymes. Furthermore, loss of outer membrane porin channels and upregulated efflux pumps do not impact the cellular uptake of cefiderocol.

• The multinational, 5-year SIDERO surveillance program reported sustained cefiderocol susceptibility rates in each year for meropenem-non-susceptible P. aeruginosa, with isolates rarely demonstrating cefiderocol minimum inhibitory concentrations >4 µg/mL.

• In two prospective, double-blind, non-inferiority clinical studies and one pathogen-focused, descriptive, open-label, clinical study cefiderocol demonstrated clinical and microbiological efficacy against P. aeruginosa in patients with complicated urinary tract infections, nosocomial pneumonia, and among critically ill patients with serious CR infections.

• Additional clinical evidence from the compassionate use program suggests that a large proportion of patients with CR P. aeruginosa, who had no alternative treatment options due to resistance, responded to cefiderocol treatment.

• Cefiderocol appears to be an appropriate choice for the treatment of CR P. aeruginosa among the current treatment options, particularly in regions where resistance to BL–BLIs has been increasingly reported.

Declaration of interest

R Canton has participated in educational programs organized by Chiesi, MSD, Pfizer and Shionogi, received research support from MSD, Shionogi, and Venatorx Pharmaceuticals, and participated in evaluation of clinical trials for MSD and Pfizer. Y Doi has received honoraria for participation in scientific advisory boards for GSK, MSD, Gilead, Shionogi, MeijiSeika, Chugai, and bioMérieux, and received speaking fees from MSD, bioMérieux, and Shionogi. P J Simner has received research grants and consulting fees from OpGen Inc. and BD Diagnostics, research grants from Affinity Biosensors and Qiagen, and consulting fees from Shionogi Inc. and GeneCapture. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

All authors contributed substantially to the conception and structure of the review article, interpretation of the relevant literature, drafting, reviewing, and editing of the article, and all authors approved the final version for submission.

Acknowledgments

Editorial and medical writing support was provided by Adrienn Kis and Joanne Shrewsbury-Gee, Highfield, Oxford, United Kingdom. This support was sponsored by Shionogi & Co., Ltd., Osaka, Japan.

Additional information

Funding

This paper was not funded.