Rapid cefiderocol NP test for detection of cefiderocol susceptibility/resistance in Enterobacterales

Abstract

Background

Cefiderocol is among the latest generation of commercialized antibiotics against a large variety of MDR Gram-negative bacteria including carbapenem-resistant Enterobacterales and non-fermenters such as Pseudomonas aeruginosa and Acinetobacter baumannii. Cefiderocol susceptibility testing, a key element for implementing rapidly a cefiderocol-based treatment, might be still challenging.

Objectives

To develop a rapid culture-based test, Rapid Cefiderocol NP test, for the identification of cefiderocol resistance among MDR Enterobacterales.

Methods

The Rapid Cefiderocol NP test is based on glucose metabolization when bacterial growth occurs and the detection of bacterial growth in the presence of cefiderocol at 64 mg/L using iron-depleted CAMHB. Bacterial growth is visually detectable by a red-to-yellow colour change of red phenol, a pH indicator. A total of 74 clinical enterobacterial isolates from various clinical sources and of worldwide origin, among which 42 isolates were cefiderocol resistant, were used to evaluate the test performance.

Results

The sensitivity and specificity of the test were found to be 98% and 91%, respectively, by comparison with the reference broth microdilution (BMD) method. All positive results were obtained within 3 h after incubation at 35°C ± 2°C, that is a gain of time of ca. 18 h (1 day) compared with currently used techniques for susceptibility testing (BMD method).

Conclusions

This novel test is rapid, highly sensitive, specific, easily interpretable, and easy to implement in routine microbiology laboratories. Such a test may rapidly and accurately provide the information needed for the implementation of adequate cefiderocol-based treatment.

Introduction

Cefiderocol is a promising antibiotic for treating infections caused by a large variety of MDR Gram-negative bacteria such as carbapenem-resistant Enterobacterales using its so-called ‘Trojan horse’ strategy.^1,2 Recently in Europe, cefiderocol was approved for treating aerobic Gram-negative-associated infections in adult patients with limited treatment options.³ It was also approved by the FDA for treating complicated urinary tract infections and hospital- and ventilator-acquired pneumonia (see the latest IDSA guidelines).⁴ In spite of the short-term and limited usage, isolates with reduced susceptibility or frank resistance to cefiderocol have been reported. Nevertheless, it seems that the potential for resistance development and acquisition remains low.^5,6 Resistance to cefiderocol is likely due to a combination of heterogeneous mechanisms such as modification of the target (PBP-3), mutations in iron transport-related proteins, co-production of serine and MBLs such as PER and NDM β-lactamases, as well as deletions, insertions and amino acid substitutions in the omega loop of AmpC enzymes as previously reported.^7–11 Moreover, several KPC mutants that are resistant to ceftazidime/avibactam (KPC-31, KPC-41, KPC-50) have reduced susceptibility to cefiderocol.^12,13

Susceptibility testing is the key element for implementing an immediate cefiderocol-based treatment. However, it might be still complex and challenging.^14,15 The gold standard method for cefiderocol susceptibility testing is MIC determination by broth microdilution method (BMD) using an iron-depleted CAMHB.¹⁶ However, this method is time-consuming and not appropriate for the routine workflow of a clinical microbiology laboratory. Other antimicrobial susceptibility testing (AST) techniques such as disc diffusion methods (30 μg cefiderocol discs), Sensititre lyophilized BMD panel (Thermo Fisher Scientific, Cleveland, OH, USA) and MIC test strip (Liofilchem, Roseto degli Abruzzi, Italy) can be alternatively used but they still require a turn-around time to get the results of at least 18 h. A recent study showed that the cefiderocol MIC strips (Liofilchem), which have been available for Pseudomonas aeruginosa only, underestimated the MICs of cefiderocol in Enterobacterales.¹⁵ Currently, clinical FDA and EUCAST and investigational CLSI breakpoints exist for interpreting cefiderocol susceptibility results for several Gram-negative bacilli including Enterobacterales (Table 1).

A rapid, early, and accurate detection of cefiderocol resistance in Enterobacterales is therefore needed. The Rapid Cefiderocol NP test was developed for the identification of cefiderocol resistance among MDR Enterobacterales. The principle of this test is based on glucose metabolization when bacterial growth occurs and the detection of bacterial growth in the presence of a given concentration of cefiderocol. Bacterial growth is visually detectable by a red-to-yellow colour change of red phenol, a pH indicator.

Methods

Bacterial strains and cefiderocol susceptibility testing

A total of 74 enterobacterial isolates obtained from the Swiss National Reference Center of Emerging Antibiotic Resistance, University of Fribourg (Switzerland), among which 42 isolates were cefiderocol resistant, were used to evaluate the performance of the Rapid Cefiderocol NP test. They were obtained from various clinical sources and of worldwide origin. The enterobacterial isolates included Escherichia coli (22 resistant, 15 susceptible), Klebsiella pneumoniae (12 resistant, 9 susceptible), Enterobacter cloacae (8 resistant, 7 susceptible) and Klebsiella aerogenes (1 susceptible). These strains had previously been characterized for their major β-lactamase content (Table 2).

Cefiderocol susceptibility testing was performed by determining MIC values with the reference BMD method using iron-depleted CAMHB, taken as a standard for the comparison with the results obtained with the Rapid Cefiderocol NP test.¹⁷ All experiments were repeated in triplicate in separate experiments. The results were interpreted according to the latest EUCAST breakpoints (https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Breakpoint_tables/v_12.0_Breakpoint_Tables.pdf).¹⁸ Isolates were categorized as susceptible when cefiderocol MICs were ≤2 mg/L and resistant when MICs were >2 mg/L. The reference strain E. coli ATCC 25922 was used as quality control for all testing.

Rapid Cefiderocol NP test

After comparison of the results with different parameters (different bacterial inocula, different cefiderocol concentrations and different incubation periods), all experiments were performed in triplicate by two different individuals with the optimal conditions obtained, as described below.

For the Rapid Cefiderocol NP test, two solutions were used, including cefiderocol stock solution and Rapid Cefiderocol NP solution. For cefiderocol stock solution, cefiderocol sulphate powder (Shionogi company) was reconstituted with sterile saline solution giving a final cefiderocol concentration of 100 mg/mL and then dispensed into small aliquots (e.g. 500 μL in 2 mL plastic vials) before being rapidly stored at −80°C for 6 months.

The Rapid Cefiderocol NP solution

Briefly, the Rapid Cefiderocol NP solution was prepared by mixing the culture medium (ID-CAMHB, 225 mL) and the pH indicator powder (0.005% of phenol red) in a glass bottle. The pH of the solution was adjusted to 7.5 (red colour) and the solution was then autoclaved at 121°C for 15 min. After cooling the solution to room temperature, we added 1% of anhydrous D(+)-glucose 10% (25 mL, Roth, Karlsruhe, Germany) sterilized by filtration. The Rapid Cefiderocol NP solution can be kept at 4°C for 1 week. This solution must be prewarmed at 37°C before use to prevent growth delay and therefore a delayed colour change and results.

Just before performing the experiment, cefiderocol was added to the Rapid Cefiderocol NP solution and mixed into sterile glass tubes to obtain the Rapid Cefiderocol NP solution containing cefiderocol at a concentration of 85.3 mg/L. As an example, 85.3 μL of cefiderocol working solution (10 mg/mL) was added to 10 mL of Rapid Cefiderocol NP solution.

Bacterial inoculum preparation

For each isolate to be tested, we prepared a standardized bacterial inoculum (0.5 McFarland standard) by resuspending freshly obtained (overnight) bacterial colonies grown on UriSelect 4 agar plates (Bio-Rad, Marnes-la-Coquette, France) or Mueller–Hinton agar plates (Bio-Rad, Marnes-la-Coquette, France) into 5 mL of sterile NaCl (0.85%). The bacterial suspensions were used within 15 min of preparation and for no longer than 1 h after preparation, as recommended by the EUCAST guidelines for susceptibility testing.

Tray inoculation

Using a 96-well polystyrene microplate (round base, with lid, sterile; Sarstedt, Germany), bacterial suspension of each isolate was inoculated in parallel into two separate wells, with and without cefiderocol. The following steps of the Rapid Cefiderocol NP test were performed, as shown in Figure 1: (1) 150 μL of cefiderocol-free Rapid Cefiderocol NP solution was transferred to wells A1–A4; (2) 150 μL of the Rapid Cefiderocol NP solution containing cefiderocol (85.3 mg/L to obtain a final concentration of 64 mg/L) was transferred to wells B1–B4; (3) 50 μL of NaCl 0.85% was added to wells A1 and B1; (4) 50 μL of the cefiderocol-susceptible isolate suspension used as negative control was added to wells A2 and B2; (5) 50 μL of the cefiderocol-resistant isolate suspension used as positive control was added to wells A3 and B3; and (6) 50 μL of the tested isolate suspension was added to wells A4 and B4. We mixed the bacterial suspension with the reactive medium by pipetting up and down (optional). The final concentration of bacteria was ≈10⁸ cfu/mL in each well, and the final concentration of cefiderocol sulfate was 64 mg/L.

Tray incubation and reading

The inoculated tray was incubated for up to 3 h at 35°C ± 2°C in ambient air without being sealed. Based on our previous experience of development of several rapid diagnostic tests,^19,20 we visually inspected the tray (checked for no spontaneous colour change) after 10 min and then every hour for 3 h. We considered the test result positive if the tested isolate grew in the presence of cefiderocol (i.e. yellow colour of the culture medium) indicating cefiderocol resistance, and negative if the tested isolate did not grow in the presence of cefiderocol (remained red) indicating no growth and, consequently, cefiderocol susceptibility.

The test result was considered interpretable in the case of the following conditions: (1) both wells with 0.85% NaCl and without bacterial suspension (A1 and B1) remained unchanged (red indicating absence of medium contamination); (2) the cefiderocol-free wells (A2–A4) with bacterial suspension turned from red to yellow, confirming the metabolism of glucose and the growth of the isolate; (3) the well (B2) with the cefiderocol-susceptible bacterial suspension (negative control) gave a negative result (remained red) confirming the absence of growth of this isolate and the susceptibility to cefiderocol; and (4) the well (B3) with the cefiderocol-resistant bacterial suspension (positive control) gave a positive result (turned yellow) confirming the viability of the isolate in the presence of cefiderocol indicating resistance to cefiderocol.

The tested isolate that grew in the absence and the presence of cefiderocol (wells A4–B4) were therefore reported to be cefiderocol resistant (Figure 1). The tested isolate that grew in the absence of but did not grow in the presence of cefiderocol were therefore reported as cefiderocol susceptible. Results were interpreted blindly by two technicians who did not know which isolates were cefiderocol resistant and cefiderocol susceptible.

Data analysis

The results obtained with the Rapid Cefiderocol NP test were compared with those obtained with the reference BMD method and discrepancies were determined for each technique. Errors [very major errors (VMEs) and major errors (MEs)] were calculated as described previously.¹⁹ An ME was considered when isolates were found to be resistant using the Rapid Cefiderocol NP test, being however susceptible by using the BMD method (false-resistant result). A VME was considered when isolates were categorized as susceptible using the Rapid Cefiderocol NP test, being however resistant by the BMD method (false-susceptible result).

Results

A total of 74 enterobacterial isolates of worldwide origin were used to evaluate the test performance. According to the results of the BMD method (Table 2), 32 were cefiderocol susceptible (MICs of cefiderocol ranging from ≤0.03 to 2 mg/L) and 42 isolates were cefiderocol resistant (MICs of cefiderocol ranging from 4 to >32 mg/L). The 32 cefiderocol-susceptible isolates gave negative results with the Rapid Cefiderocol NP test, except for three isolates (MICs of 1 to 2 mg/L), among which two isolates had borderline susceptibility with cefiderocol MICs of 2 mg/L, which gave positive (false-positive) results (Table 2). The cefiderocol-resistant isolates (n = 42) were all correctly identified as cefiderocol resistant by using the Rapid Cefiderocol NP test except for one isolate (cefiderocol MIC value of 4 mg/L), which gave a negative (false-negative) result (Table 2). Overall, three MEs (false resistance) but one VME (false susceptibility) were observed. Consequently, strong agreement was found between the results of the BMD susceptibility testing and the Rapid Cefiderocol NP test results (both for susceptible and resistant isolates). The sensitivity and specificity of the test were calculated to be 98% (95% CI 87.7%–99.9%) and 91% (95% CI 75.8%–96.8%), respectively, by comparison with the BMD method taken as the reference standard method, with an overall performance of 95% categorical agreement, 9.4% ME rate (3 of 32 cefiderocol-susceptible isolates) and 2.4% VME rate (1 of 42 cefiderocol-resistant isolates). By reading the colour change of the wells every hour, we determined that final results shall be best read at 3 h after incubation at 35°C ± 2°C under an ambient atmosphere.

Discussion

Due to major and continuing emergence of MDR Gram negative bacteria, the place of novel antibiotics such as cefiderocol is obvious. However, the susceptibility testing for cefiderocol, the key for the success of a cefiderocol-based treatment, is still challenging.²¹ EUCAST guidelines make the reading of BMD cefiderocol results difficult and subject to various interpretations depending on the consciousness and expertise of the reader leading under- or over-estimation of cefiderocol MIC values.¹⁵ To date, an iron-depleted CAMHB BMD, a lyophilized BMD panel and disc diffusion methods are available for antimicrobial susceptibility testing of cefiderocol. However, those methods are time-consuming (18–24 h). On the other side, development of MIC test strips and automated cefiderocol susceptibility testing panels, which still remains difficult, will facilitate clinical laboratory efforts.²¹

Here, a test for rapid testing of cefiderocol susceptibility has been settled, combining good sensitivity and specificity. This test was able to detect cefiderocol resistance among Enterobacterales, regardless of their resistance mechanisms, within 3 h. Such a time frame is at least 16–18 h earlier (meaning 1 day earlier from a clinical point of view) than the reference BMD method. Even though a very few discrepancies were observed (three MEs and one VME), the sensitivity and specificity of the Rapid Cefiderocol NP test were high, making it a potential useful AST tool for clinical laboratories. Moreover, this test is easy to implement in routine microbiology laboratories requiring a single methodological step. The Rapid Cefiderocol NP test is also easily interpretable with a very rapid susceptibility/resistance categorization, which is the information needed for the implementation of an adequate cefiderocol-based treatment. The Rapid Cefiderocol NP test requires the preparation and use of an iron-depleted CAMHB, which could be cumbersome and time-consuming; however, this test would be suitable for any microbiology laboratory, in particular if a commercial version is developed. An additional limitation point is that this test could not be applied for the CLSI and FDA breakpoints. However, modifications of the cefiderocol concentration of the test may lead to a test fitted to the CLSI and FDA breakpoints.

Conclusions

Since long turn-around time and a great deal of time are spent per patient for coordinating AST for such novel agents (cefiderocol),²² a newly developed approach such as the Rapid Cefiderocol NP fits with what is expected from a rapid diagnostic test (besides accuracy with high sensitivity and specificity), namely optimal rapidity and easiness. In addition, it does not need any additional equipment for its reading. Clinical evaluation of the proposed test may be interesting, as well as further development of a test for detection of cefiderocol susceptibility/resistance among MDR Acinetobacter baumannii and P. aeruginosa. Owing to the size of our collection, which may be considered as a limitation, further validation of this test in different laboratories with larger and more diverse sets of cefiderocol-resistant and -susceptible enterobacterial isolates will be needed.

Acknowledgements

Cefiderocol powder was provided by the manufacturer (Shionogi & Co., Ltd.). This work was partially presented as a poster (P0986) at the 32nd European Congress of Clinical Microbiology and Infectious Diseases (ECCMID 2022), April 2022, Lisbon, Portugal.

Funding

This work has been funded by the University of Fribourg, the Swiss National Science Foundation (project FNS-407240_177381).

Transparency declarations

None to declare.

References