https://orcid.org/0000-0002-2630-9722
A recent study in Journal of Antimicrobial Chemotherapy1 reported high levels of cefiderocol resistance in carbapenemase-producing Enterobacterales (CPE) using EUCAST disc diffusion breakpoints at a university teaching hospital in London, UK.
The study reported use of the disc diffusion method with 30 μg cefiderocol discs (Liofilchem, Abruzzi, Italy) on standardized supplemented Mueller–Hinton agar plates.1 Using EUCAST breakpoints (version 14.0), 113 CPE isolates were evaluated for cefiderocol susceptibility. Of these isolates, 41/113 (36%) were reported as susceptible (zone size ≥ 23 mm), with 9/41 (22%) of these isolates within the area of technical uncertainty (ATU, 21–23 mm). Cefiderocol resistance (zone size < 23 mm) was reported in 72/113 (64%) isolates with 17/72 (24%) within the ATU.1
EUCAST currently recommend that laboratories initially perform cefiderocol disc diffusion on isolates due to its reliability outside of the ATU in predicting susceptibility and resistance.2 Inside the ATU, if there are no alternative methods to determine susceptibility, such as broth microdilution or reference laboratory assessment, they recommend interpreting as per the zone diameter.2 There have been several warnings issued by EUCAST regarding the challenges of cefiderocol susceptibility testing, with several studies suggesting that disc diffusion may require supplementation with broth microdilution, especially for isolates within the ATU.2–6 The challenge for laboratories is that production and quality control of iron-depleted CAMHB can be difficult. Commercial testing kits using the gradient strip method, such as MIC gradient test strips by Liofilchem (USA), have failed to demonstrate essential agreement compared with gold-standard broth microdilution.7 Commercial broth microdilution kits (ComASP) for cefiderocol susceptibility testing are available and have shown high levels of agreement with disc diffusion,3,6 but cannot be used as a reference method.2,8
At Imperial College Healthcare NHS Trust, high rates of cefiderocol resistance have been observed using disc diffusion alone despite EUCAST quality control requirements being met. To explore this further, cefiderocol susceptibility testing using a commercial broth microdilution method (ComASP) in line with EUCAST guidance was implemented.8 Using broth microdilution to determine MIC, our verification suggests that cefiderocol disc diffusion may not be reliable to confirm resistance in Enterobacterales (Figure 1).
Comparison of disc diffusion with cation-adjusted broth microdilution results for isolates. Nil indicates no acquired β-lactamase enzyme detected using conventional phenotypic and genotypic laboratory diagnostics. Asterisks indicate results within the ATU. Red, resistant; green, susceptible.
Disc diffusion quality control is performed daily using Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853. Disc diffusion performs well with the mean of at least five repeated tests within ±1 mm of the target zone size (27 mm Enterobacterales and 26 mm P. aeruginosa).2,9 For broth microdilution, triplicate repeats were performed using E. coli ATCC 25922 and P. aeruginosa ATCC 27853. Mean MIC was within the target range for both isolates performing tandem repeats with mean MICs of 0.125 and 0.25 mg/L, respectively.8,9
As part of internal verification of cefiderocol broth microdilution, broth microdilution of 54 clinical isolates including Enterobacterales (45/54; 83%), P. aeruginosa (6/54; 11%) and Acinetobacter baumannii (3/54; 6%) with variable phenotypes was compared with disc diffusion (Figure 1). For disc diffusion, zone diameters (mm) were interpreted using EUCAST breakpoints (version 14.0). Broth microdilution was performed following the manufacturer’s recommendations and MIC determined visually in line with EUCAST guidelines.8
Using disc diffusion, isolates were classified as resistant to cefiderocol if zone sizes were <23 mm for Enterobacterales and <22 mm for P. aeruginosa. For broth microdilution, isolates were deemed resistant if the MICs were >2 mg/L. Although no breakpoints are available for A. baumannii, zone sizes of ≥17 mm have been demonstrated to correspond to a pharmacokinetic/pharmacodynamic breakpoint of MIC ≤ 2 mg/L (EUCAST breakpoints 14.0).
No isolates reported as susceptible on disc diffusion were identified as resistant using broth microdilution for any species evaluated within this verification. For Enterobacterales, 24/45 (53%) isolates were classified as resistant using disc diffusion. Five of the resistant isolates (21%) fell within the ATU. Using broth microdilution, 5/45 (11%) were classified as cefiderocol resistant. For P. aeruginosa, 2/6 (33%) of isolates were classified as resistant using disc diffusion with both falling within the ATU. Using broth microdilution, 1/6 (17%) were classified as resistant. For A. baumannii, both disc diffusion and broth microdilution classified 2/3 (67%) of isolates resistant.
These results demonstrate no very major errors (false susceptibility), but do demonstrate major errors (false resistance) in 20/28 (71%) cases of reported cefiderocol resistance using the disc diffusion method. These findings have led to our laboratory routinely determining cefiderocol MIC on any resistant isolates identified by the disc diffusion method. Genome sequencing data is not currently available to confirm isolate genotype but is planned to explore whether potential genotypic variations are contributing to these observed differences.
We believe that these findings highlight the ongoing challenge of cefiderocol susceptibility testing as noted by EUCAST. Reliance on disc diffusion results alone may overcall cefiderocol resistance when compared with broth microdilution and may lead to patients receiving suboptimal treatment that puts them at increased risk of treatment failure or toxicity. Our experience suggests that cefiderocol resistance determined using disc diffusion should be confirmed using broth microdilution, regardless of zone size.
Funding
No funding was received to support this work.
Transparency declarations
T.M.R. has received honoraria from Sandoz (2020), bioMérieux (2021–25), Pfizer (2024) and Roche Diagnostics (2021). All other authors have no conflicts of interest to declare.
