Cefiderocol AST in a real-life Klebsiella pneumoniae collection: challenges in the ATU range.

Objectives: FDC susceptibility testing is challenging as none of the commercial tests have been proven to accurately determine the susceptibility in the area of technical uncertainty (ATU). Here, we evaluated the performance of different FDC testing methods on Klebsiella pneumoniae isolates around this range.

Methods: A challenging collection of 104 K. pneumoniae isolates with different FDC susceptibility was designed, including high representation (70%) of the 2024-ATU (21-23 mm) ± 1 mm. MICs were determined by broth-microdilution (BMD) reference method, commercial BMD (ComASP®, EUMDROXF®), MIC-gradient strips (Liofilchem) and disk-diffusion (DD) (Liofilchem, ThermoFisher). MIC devices evaluation was performed following ISO 20776-2:2021, calculating essential agreement (EA) and bias. DD results were evaluated following 2023 and 2024-EUCAST-guidelines, calculating major errors (ME) and very major errors. Categorical agreement (CA) was determined for all the methods.

Results: Overall, EUMDROXF® and ComASP® showed 81.7% (95% CI = 72.9-88.6) and 88.5% (95% CI = 80.8-93.9) EA, +2.6% and +27.9% bias and 99.0% (95% CI = 94.7-99.9) and 98.1% (95% CI = 93.3-99.8) CA, respectively. Liofilchem MIC-gradient strips exhibited 47.1% (95% CI = 37.2-57.1) EA, +2.9% bias and 93.3% (95% CI = 86.7-97.3) CA. In DD, variability between manufacturers was elevated. CA lowered and ME increased more than 10% with 2024-EUCAST-breakpoints modification.

Conclusions: DD performance was insufficient to assess FDC resistance in K. pneumoniae and modification of EUCAST-breakpoints did not solve the problem. ComASP® panel fulfilled ISO criteria and could be used as MIC-confirmatory method, at least in K. pneumoniae. However, EUMDROXF®, even close, did not fulfil the EA criterion. MIC-gradient strips exhibited major limitations.