Comparison of disk diffusion, MIC test strip and broth microdilution methods for cefiderocol susceptibility testing on carbapenem-resistant enterobacterales.

ABSTRACT

OBJECTIVES: Cefiderocol is a catechol-substituted cephalosporin dedicated to the treatment of infections caused by multidrug resistant gram-negative rods. Cefiderocol susceptibility testing might be complex. We compared cefiderocol susceptibility testing methods on a relevant collection of carbapenem-resistant Enterobacterales. METHODS: CE-IVD (European CE marking required for all in vitro diagnostic (IVD)) broth microdilution (BMD) plate (ThermoFisher, Waltham, MA, USA) using regular Mueller-Hinton broth, MIC test strip (Liofilchem, Teramo, Italy), and disk diffusion (Liofilchem) were compared to a frozen BMD plate prepared with iron depleted Mueller-Hinton broth. First, a collection of 100 entirely sequenced carbapenem-resistant Enterobacterales was used to compare these methods. Then, a prospective comparison of disk diffusion and CE-IVD BMD was performed on 827 consecutive carbapenem non-susceptible Enterobacterales including 634 carbapenemase-producers. RESULTS: Compared to reference method, CE-IVD BMD plate gave 95.0% (95% CI, 88.8-97.9) categorisation agreement (CA), 2.8% (95% CI, 0.4-14.2) very major errors (VME), and 1.6% (95% CI, 0.3-8.7) major errors (ME) with high reproducibility. MIC strip gave only 63% (95% CI, 53.2-71.8) of CA and 94.9% (95% CI, 83.1-98.6) of VME due to critical underestimation of the MICs. Disk diffusion gave 77% (95% CI, 67.9-84.2) CA with additional 8% of the isolates within the area of technical uncertainty of 18-22 mm. Prospectively, disk diffusion gave 81.7% (95% CI, 79.0-84.2) CA, 23.3% (95% CI, 15.1-34.2%)VME, and 4.9% (95% CI, 3.6-6.7) ME. Additionally, 21.3% (95% CI, 18.6-24.2) of CRE were within the area of technical uncertainty. DISCUSSION: Commercial CE-IVD BMD (ThermoFisher) is accurate for cefiderocol MIC determination in difficult-to-treat Enterobacterales whereas MIC test strip (Liofilchem), that was formulated for Pseudomonas aeruginosa only, should be avoided. Disk diffusion might be useful for screening, but many of these CRE have to be re-tested using BMD to assess definitive categorization.