Burkholderia pseudomallei Clinical Isolates Are Highly Susceptible In Vitro to Cefiderocol, a Siderophore Cephalosporin

Cefiderocol is a cephalosporin designed to treat multidrug-resistant Gram-negative infections. By forming a chelated complex with ferric iron, cefiderocol is transported into the periplasmic space via bacterial iron transport systems and primarily binds to penicillin-binding protein 3 (PBP3) to inhibit peptidoglycan synthesis. This mode of action results in cefiderocol having greater in vitro activity against many Gram-negative bacilli than currently used carbapenems, β-lactam/β-lactamase inhibitor combinations, and cephalosporins.

ABSTRACT

Cefiderocol is a cephalosporin designed to treat multidrug-resistant Gram-negative infections. By forming a chelated complex with ferric iron, cefiderocol is transported into the periplasmic space via bacterial iron transport systems and primarily binds to penicillin-binding protein 3 (PBP3) to inhibit peptidoglycan synthesis. This mode of action results in cefiderocol having greater in vitro activity against many Gram-negative bacilli than currently used carbapenems, β-lactam/β-lactamase inhibitor combinations, and cephalosporins. Thus, we investigated the in vitro activity of cefiderocol against a total of 246 clinical isolates of Burkholderia pseudomallei from Queensland, Australia. The collection was composed primarily of bloodstream (56.1%), skin and soft tissue (16.3%), and respiratory (15.9%) isolates. MICs of cefiderocol ranged from ≤0.03 to 16 mg/liter, whereas the MIC₉₀ was 0.125 mg/liter. Based upon CLSI clinical breakpoints for cefiderocol against Pseudomonas aeruginosa, Acinetobacter baumannii, and Stenotrophomonas maltophilia, three isolates (1.2%) would be classified as nonsusceptible (MIC > 4 mg/liter). Using EUCAST non-species-specific (pharmacokinetic/pharmacodynamic [PK/PD]) clinical breakpoints or those set for Pseudomonas aeruginosa, four isolates (1.6%) would be resistant (MIC > 2 mg/liter). Further testing for coresistance to meropenem, ceftazidime, trimethoprim-sulfamethoxazole, amoxicillin-clavulanate, and doxycycline was performed on the four isolates with elevated cefiderocol MICs (>2 mg/liter); all isolates exhibited resistance to amoxicillin-clavulanic acid, while three isolates also displayed resistance to at least one other antimicrobial. Cefiderocol was found to be highly active in vitro against B. pseudomallei primary clinical isolates. This compound shows great potential for the treatment of melioidosis in countries of endemicity and should be explored further.

INTRODUCTION

Burkholderia pseudomallei is endemic to tropical and subtropical regions, including Southeast Asia and northern Australia, and has a presence in Africa and South America (1–3). The bacteria are known to cause melioidosis, where pneumonia, sepsis, neurological disease, and visceral abscesses are commonly described clinical presentations (1, 4). Treatment for B. pseudomallei infections requires an intensive 2- to 8-week intravenous treatment with an antimicrobial such as a carbapenem or cephalosporin, followed by an orally administered eradication treatment for 3 to 6 months with an antimicrobial such as trimethoprim-sulfamethoxazole (5). Infections frequently require intensive care admission and are associated with significant morbidity (6). The associated morbidity is due, in part, to the intrinsic antimicrobial resistance (AMR) of B. pseudomallei as well as acquired resistance to antimicrobials such as tetracyclines, β-lactam/β-lactamase inhibitors, and, rarely, carbapenems (7–9). The mortality rate for patients presenting with melioidosis ranges from 10% to 40% in Burma, Singapore, Thailand, and Vietnam (6), while in Australia, it remains approximately 10% (10). Recent research has focused on new or repurposed compounds in order to provide improved treatment options for such infections, particularly given the potential of B. pseudomallei to be genetically manipulated or used as an agent of biologic warfare (11, 12).

Cefiderocol (formerly S-649266; Shionogi & Co. Ltd., Osaka, Japan) inhibits peptidoglycan synthesis and has been described as almost ubiquitously stable against β-lactamases, resulting in greater efficacy than those of carbapenems, currently available β-lactam/β-lactamase inhibitor combinations, and cephalosporins (13–16). It has great potential for the treatment of pathogenic multidrug-resistant (MDR) and carbapenem-nonsusceptible Gram-negative bacilli (GNB), as they have increased significantly worldwide (17, 18). This has resulted in diminished appropriate antimicrobial therapy options (19, 20).

Cefiderocol recently received FDA approval for the treatment of complicated urinary tract infections (21) on the basis of noninferiority against imipenem/cilastatin (22) as well as approval for the treatment of hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia (23, 24). Another trial against nosocomial pneumonia has also been completed (13, 25), with efficacy demonstrated against Gram-negative pathogens such as Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii (26, 27). Provisional CLSI (Clinical and Laboratory Standards Institute) in vitro breakpoints have been set at ≤4 mg/liter (susceptible [S]) and ≥16 mg/liter (resistant [R]) for these pathogens (26). Recently, EUCAST (European Committee on Antimicrobial Susceptibility Testing) set much lower in vitro breakpoints of ≤2 mg/liter (susceptible) and >2 mg/liter (resistant) for Enterobacterales and Pseudomonas aeruginosa and as a pharmacokinetic/pharmacodynamic (PK/PD; non-species specific) breakpoint (28). Furthermore, cefiderocol has shown promising efficacy against Burkholderia pseudomallei and Burkholderia mallei, resulting in MIC₉₀s of 0.25 and 4 mg/liter, respectively (n = 30) (29). Therefore, with the need for new treatment options for B. pseudomallei or other GNB and the promising efficacy of cefiderocol as a treatment option, the aim of this study was to assess the in vitro activity of cefiderocol against a large sample size of clinical B. pseudomallei isolates from Queensland, Australia, where melioidosis is endemic.

RESULTS

Disc diffusion.

All 246 isolates were subjected to cefiderocol disc diffusion testing. Zone diameters ranged from 11 to 46 mm (Table 1). One isolate, C137 (38 mm), demonstrated inner growth on cefiderocol disc diffusion testing (12 mm), indicating a heterogeneous population or expression. The resistant subpopulation, deemed C137R, was directly subjected to broth microdilution (BMD) susceptibility testing. However, the resistant subpopulation could not be isolated via subculture from the C137 sample; therefore, the original C137 sample was used for further testing.

TABLE 1.

In vitro MICs and disc diffusion ranges for clinical B. pseudomallei isolates against cefiderocol

Species	No. of isolates at a cefiderocol MIC (mg/liter) of:											MIC (mg/liter)			Disc zone (mm)
	≤0.03	0.06	0.12	0.25	0.5	1	2	4	8	16	≥32	Range	50%	90%
B. pseudomallei (n = 246)	112	89	37	1	2	1	0	1	1	2	0	≤0.03–16	0.06	0.125	11–46

Broth microdilution.

Cefiderocol BMD testing was performed on all 246 isolates. The MIC range, MIC₅₀, and MIC₉₀ were ≤0.03 to 16, 0.06, and 0.125 mg/liter, respectively (Table 1; Fig. 1). An association between zone diameter and MIC was only observed for MIC values ≥8 mg/liter (Fig. 2).

FIG 1.

Histogram of cefiderocol MICs (mg/liter) for 246 clinical Burkholderia pseudomallei isolates.

FIG 2.

Scatterplot of cefiderocol disc zone diameters (mm) versus cefiderocol MICs (mg/liter) for 246 clinical Burkholderia pseudomallei isolates.

Isolates expressing elevated cefiderocol MICs.

According to CLSI breakpoints for cefiderocol against other GNB, two isolates (0.8%) would be categorized as resistant (MIC, ≥16 mg/liter), while one isolate (0.4%) would be categorized as intermediate (MIC, 8 mg/liter). With respect to EUCAST breakpoints, four isolates would be deemed resistant (MIC, >2 mg/liter). These four isolates comprised two lung specimens, one from a cystic fibrosis patient and one from a persistent B. pseudomallei infection, one skin and soft tissue infection, and one bloodstream infection. All four individuals had serious underlying comorbidities. Coresistance was examined via BMD testing between cefiderocol and meropenem, ceftazidime, amoxicillin-clavulanic acid, doxycycline, and trimethoprim-sulfamethoxazole; all isolates exhibited resistance to amoxicillin-clavulanic acid, while three isolates also displayed resistance to meropenem or trimethoprim-sulfamethoxazole (Table 2).

TABLE 2.

MICs of B. pseudomallei clinical isolates with elevated cefiderocol concentrations

Isolate IDa	B. pseudomallei isolate metadata collection		Clinically relevant antimicrobial (mg/liter) and categoryb
	Site	Yr	FDC	MEM	CAZ	AMC	SXT	DOX
C137Rc	SSTI	2016	16	1	2	>32 R	0.5/9.5	2
C4	Lung	2006	8	>16 R	2	>32 R	0.12/2.28	2
C50	Lung	2000	4	4	≥16 NS	>32 R	4/76 R	4
C194	Blood	2018	16	>16 R	2	>32 R	0.12/2.28	2

^aID, identifier.

^bFDC, cefiderocol. Clinically relevant antimicrobials included meropenem (MEM), ceftazidime (CAZ), amoxicillin-clavulanic acid (AMC), trimethoprim-sulfamethoxazole (SXT), and doxycycline (DOX). Nonsusceptible (NS) and resistant (R) categories were determined by CLSI breakpoints.

^cBroth microdilution testing was performed on isolate C137 as the resistant subpopulation was not able to be subcultured from this sample.

DISCUSSION

B. pseudomallei infections pose a significant mortality risk to patients. Furthermore, treatment can be especially burdensome in terms of the need for prolonged intravenous and oral antibiotic therapy. Due to the virulence of the organism and its predilection for immunocompromised hosts, patients with melioidosis frequently present with severe disease requiring intensive care admission. While uncommon, emergent resistance may also occur to compromise treatment. A compound such as cefiderocol with such low MICs could be used as an empirical therapy to improve clinical outcomes of B. pseudomallei infection in regions of endemicity, where multidrug-resistant Gram-negative bacilli are circulating. An approach such as this may result in improved clinical responses during empirical therapy and increased bacterial killing in difficult-to-treat infections, such as those in the central nervous system, bone, joint, and lymphatic system.

Disc diffusion.

The relationship between cefiderocol disc diffusion zone diameter and BMD MIC values was observed at concentrations ≥8 mg/liter of cefiderocol. As there are no zone diameters specified for cefiderocol disc diffusion against B. pseudomallei, very major errors, major errors, and minor errors were unable to be calculated. Zone diameter was able to differentiate isolates with MICs of ≥8 mg/liter consistently at 25 mm or less. Isolates with MICs between 1 and 4 mg/liter ranged between 35 to 25 mm, where the relationship may be more prominent with a larger number of isolates with MICs within this range (Fig. 2). This relationship was previously observed in more than 1,300 Gram-negative bacteria, where isolates with MICs of ≥8 mg/liter were consistently differentiated at 20 mm or less (30). These relationships were present in species within the Enterobacterales as well as P. aeruginosa and A. baumannii; however, Stenotrophomonas maltophilia isolates (≥2 mg/liter) were differentiated at 15 mm or less (30). Better discrimination among susceptible isolates may be seen with a lower disc mass, but only 30-μg discs are currently available commercially.

Broth microdilution.

In this study, isolates with MIC values of >2 mg/liter were considered to be elevated based on EUCAST clinical breakpoints set for Enterobacterales, P. aeruginosa, and S. maltophilia (31). Only a few were identified in this collection (n = 4; 1.6%), coinciding with the rates previously reported for Australian B. pseudomallei (0.0% to 4.0%) against other clinically relevant antimicrobials (32–34). This finding is expected, with previous in vitro studies identifying very few resistant isolates for several Gram-negative bacilli species. For example, only 24/753 (3.2%) isolates showed MICs of ≥8 mg/liter from a global data set (35), while 54/8,954 (0.6%) isolates in North America and Europe had MICs of ≥4 mg/liter (27), and no resistant isolates were identified among 189 isolates from Greece (36). The MIC₉₀ of the 246 B. pseudomallei isolates in this study was 0.125 mg/liter, not significantly less than that of the previously described MIC₉₀ of 0.25 mg/liter obtained from a study of 30 B. pseudomallei isolates (29). However, an MIC₉₀ of 0.125 mg/liter is similar to that observed in the closely related Burkholderia cepacia complex, with MIC₉₀ values of 0.016 to 0.12 mg/liter attained from small studies (13, 27), yet much less than that of sister species B. mallei, with an MIC₉₀ of 4 mg/liter (29). Notably, an increase in sample size has not substantially increased the MIC₉₀ of cefiderocol in other pathogens such as P. aeruginosa, A. baumannii, K. pneumoniae, and Escherichia coli (27, 35, 36). Therefore, it is reasonable to assume that the MIC₉₀ of Australian B. pseudomallei will remain low against cefiderocol, even with a dramatic increase in sample size.

Isolates with elevated cefiderocol MICs.

As no individual had received cefiderocol previously, we suggest any MIC increase is likely a reflection of including isolates from clinically complex individuals with underlying comorbidities. The four isolates with MICs of ≥4 mg/liter were derived from individuals with cystic fibrosis, diabetes mellitus, and persistent B. pseudomallei infection. Our results predict elevated cefiderocol MICs will more than likely be encountered in individuals that may have previously received extensive antimicrobial exposure due to underlying comorbidities such as cystic fibrosis or in the case of persistent or relapsing B. pseudomallei infections.

Five clinically relevant antimicrobials were selected for further BMD testing against B. pseudomallei isolates with elevated cefiderocol MICs to assess the presence of coresistance. Interestingly, all four isolates exhibited resistance to amoxicillin-clavulanic acid (introduced by penA mutations in B. pseudomallei [8]), while one of these isolates also exhibited nonsusceptibility to ceftazidime and resistance to trimethoprim-sulfamethoxazole. Two isolates also exhibited meropenem resistance. Previous studies have highlighted that cefiderocol activity is not generally impacted by resistance to other antimicrobials (13–15, 27, 36). Our findings agree, with amoxicillin-clavulanic acid, meropenem, ceftazidime, or trimethoprim-sulfamethoxazole resistance not appearing to impact cefiderocol MICs. As these isolates are from individuals with known significant comorbidities and who were likely exposed to antimicrobials previously, it is suggested that an increase in MIC could be a result of exposure to multiple antimicrobials in these individuals. Characterization of the mechanism of resistance in strains with elevated MIC is under way.

Prospective role of cefiderocol against B. pseudomallei infections.

Cefiderocol has great potential as an antimicrobial therapy for multidrug-resistant B. pseudomallei infections. This is supported by the findings of this study in combination with the efficacy of cefiderocol being unaffected by resistance to other antimicrobials in vitro (36) as well as the antimicrobial being well tolerated by patients (13, 16) and found to be safe in clinical trials to date (16). However, this study also suggests that cefiderocol may lose efficacy in vitro against B. pseudomallei isolates derived from individuals with significant underlying comorbidities or with significant prior antibiotic exposure. Nevertheless, further in vitro testing, followed by in vivo trials of cefiderocol against B. pseudomallei, is warranted.

Conclusion.

Cefiderocol demonstrates a high degree of activity in vitro against 246 clinical isolates of B. pseudomallei from Queensland, Australia. The MIC₅₀ and MIC₉₀ were 0.06 and 0.125 mg/liter, respectively. Resistant isolates with MICs of >2 mg/liter were infrequently demonstrated in this collection (1.6%) based on EUCAST non-species-specific breakpoints and were most commonly associated with significant underlying comorbidities. Cefiderocol shows promise as an intravenous agent for the management of acute melioidosis based upon in vitro susceptibility testing, particularly in regions where carbapenem-resistant Gram-negative organisms and B. pseudomallei cocirculate. Further investigation into the role of cefiderocol as a treatment for melioidosis and likely mechanisms of resistance would be of great value.

MATERIALS AND METHODS

Isolate collection and storage.

Clinical B. pseudomallei isolates were prospectively collected from patients admitted to Queensland Health hospitals, Australia, over a period from 1999 to 2018. All isolates during this period were glycerol stocked at the time of collection and stored at −80°C until further use. A small number of isolates were referred from external laboratories. These isolates were retrieved from −80°C storage from three microbiology laboratories in Queensland (Forensic and Scientific Services, Coopers Plains, Australia; Pathology Queensland and Townsville and Central laboratories). Isolates were transferred to the University of Queensland Centre for Clinical Research (UQCCR) and stored at −80°C prior to testing. Demographic and clinical information for the isolates and patients was retrieved from the laboratory information system (Auslab; PJA Solutions). Ethical approval for this study was granted by the Forensic and Scientific Services Human Ethics Committee (HREC/17/QFSS/12). Biosafety approvals for this study were granted by the Institutional Biosafety Committee UQCCR (IBC/210B/SOM/2017). This project was performed under the study number S-649266-EF-312-N.

Isolates.

A total of 246 isolates from 246 individuals were included in this study (Table 3). All isolates were from Queensland, where B. pseudomallei was predominantly isolated from the blood, lung, skin, and soft tissues of male patients (Table 3).

TABLE 3.

Demographic and clinical data for primary B. pseudomallei isolates

Category	Value(s)
Demographics
No. of individuals	246
Age range (yrs)	6–90
Gender (%)
Female	25.6
Male	73.2
Unknown	1.2
Clinical data
Isolate collection site (n)
Blood	138
Skin and soft tissue	40
Sputum	21
Lung	18
Urine	12
Bone/joint/fluid	4
Unknown	3
Pleural fluid	3
Liver	1
Prostate	1
Brain	1
Lymph node	1
Gastrointestinal tract	1
Gut	1
Peritoneal fluid	1
No. of isolates	246

Disc diffusion.

Disc diffusion susceptibility testing was performed using 30-μg cefiderocol discs provided by Mast Group Ltd. (Bootle, UK). Discs were stored at 4°C prior to use. Isolates were subcultured from storage on 5% horse blood agar (Micromedia, Victoria, Australia) for 18 to 24 h at 37°C prior to preparation of a 0.5 McFarland solution in 0.9% sterile saline. Mueller-Hinton agar plates (Micromedia) were inoculated with test isolates using the Kirby-Bauer method and incubated for 16 to 20 h at 37°C. Control strains E. coli ATCC 25922 and P. aeruginosa ATCC 27853 were included with each run (Table 4).

TABLE 4.

Accepted range of MICs and disc inhibition zone diameters for cefiderocol against quality control organisms E. coli and P. aeruginosa and respective non-species-specific breakpoints

Organism	MIC (mg/liter)				Zone diameter (mm)
	CLSI		EUCAST		CLSI	EUCAST
	Range	Break pointa	Range	Break pointb
E. coli ATCC 25922	0.06–0.5		0.06–0.5		25–31	24–30
P. aeruginosa ATCC 27853	0.06–0.5		0.06–0.5		22–31	22–28
Non-species specific
Susceptible		≤4		≤2
Resistant		≥16		>2

^aCLSI breakpoints set for Enterobacterales, P. aeruginosa, A. baumannii, and S. maltophilia.

^bEUCAST PK/PD, Enterobacterales and P. aeruginosa breakpoints.

Broth microdilution.

Broth microdilution (BMD) testing was performed using 96-well plates provided by Shionogi & Co., Ltd. (Osaka, Japan) and prepared by International Health Management Associates (IHMA; Schaumburg, USA). Iron-depleted cation-adjusted Mueller-Hinton broth (ID-CAMHB) was used according to CLSI and the manufacturer’s recommendations (37). Plates were stored at −20°C and thawed for 1 h prior to use. Isolates were subcultured from storage on to 5% horse blood agar (Micromedia, Victoria, Australia) for 18 to 24 h at 37°C prior to preparation of a 0.5 McFarland solution in 0.85% sterile saline for a final inoculum of 5 × 10⁵ cells, as per CLSI BMD guidelines. Plates were inoculated once for each isolate and incubated at 37°C for 16 to 20 h prior to reading. One positive-growth-control well and one negative-control well were included in each plate. Quality control of cefiderocol was performed using E. coli ATCC 25922 and P. aeruginosa ATCC 27853 (Table 4) (38). CLSI provides susceptible (≤4 mg/liter) and resistant (≥16 mg/liter) interpretative criteria for Enterobacterales and some nonfermenters such as Pseudomonas aeruginosa, Acinetobacter baumannii, and Stenotrophomonas maltophilia (but not B. pseudomallei) (26). More recently, EUCAST breakpoints of susceptible (≤2 mg/liter) and resistant (>2 mg/liter) have been described (28), and as such, EUCAST breakpoints were applied to isolates in this study (Table 4).

Further antimicrobial susceptibility testing was performed on isolates with cefiderocol MICs of >2 mg/liter (n = 4), including clinically relevant antimicrobials meropenem, ceftazidime, doxycycline, amoxicillin-clavulanic acid, and trimethoprim-sulfamethoxazole (Table 2). BMD testing was performed in-house where 96-well plates were prepared using the Tecan D300e Digital Dispenser (HP Inc., CA, USA). Standard inoculation (5 × 10⁵ CFU), incubation in ambient air at 37°C, and endpoint reading at 24 h were performed. Quality control was performed in duplicates for each batch of plates made, including E. coli ATCC 25922, E. coli ATCC 35218, and S. aureus ATCC 29213, to ensure correct dilutions of each antimicrobial (38). The ranges of meropenem, ceftazidime, doxycycline, amoxicillin-clavulanic acid, and trimethoprim-sulfamethoxazole tested were 0.06 to 16, 0.06 to 16, 0.12 to 16, 0.12/2 to 32/2, and 0.06/1.19 to 32/608 mg/liter, respectively. MICs were interpreted according to CLSI breakpoints (39).

Longitudinal analysis.

Multiple isolates collected from 11 patients over time were also identified in the collection (n = 13). These isolates comprised specimens collected from an alternate anatomical site or isolates collected weeks, months, or years apart. The antimicrobial treatment of patients remains unknown. These isolates were subject to the same methods as described above and are presented in Tables S1 and S2 the supplemental material.