Abstract
Antimicrobial susceptibility testing (AST) of cefiderocol poses challenges because of its unique mechanism of action (i.e., requiring an iron-depleted state) and due to differences in interpretative criteria established by the Clinical and Laboratory Standards Institute (CLSI), U.S. Food and Drug Administration (FDA), and European Committee on Antimicrobial Susceptibility Testing (EUCAST). Our objective was to compare cefiderocol disk diffusion methods (DD) to broth microdilution (BMD) for AST of Gram-negative bacilli (GNB). Cefiderocol AST was performed on consecutive carbapenem-resistant Enterobacterales (CRE; 58 isolates) and non-glucose-fermenting GNB (50 isolates) by BMD (lyophilized panels; Sensititre; Thermo Fisher) and DD (30 μg; research-use-only [RUO] MASTDISCS and FDA-cleared HardyDisks). Results were interpreted using FDA (prior to 28 September 2020 update), EUCAST, and investigational CLSI breakpoints (BPs). Categorical agreement (CA), minor errors (mE), major errors (ME), and very major errors (VME) were calculated for DD methods. The susceptibilities of all isolates by BMD were 72% (FDA), 75% (EUCAST) and 90% (CLSI). For DD methods, EUCAST BPs demonstrated lower susceptibility at 65% and 66%, compared to 74% and 72% (FDA) and 87% and 89% (CLSI) by HardyDisks and MASTDISCS, respectively. CA ranged from 75% to 90%, with 8 to 25% mE, 0 to 19% ME, and 0 to 20% VME and varied based on disk, GNB, and BPs evaluated. Both DD methods performed poorly for Acinetobacter baumannii complex. There is considerable variability when cefiderocol ASTs are interpreted using CLSI, FDA, and EUCAST breakpoints. DD offers a convenient alternative approach to BMD methods for cefiderocol AST, with the exception of A. baumannii complex isolates.